AI Safety Compendium

❯

Folder: summaries

618 items under this folder.

29 Apr 2026
AI Safety Atlas Ch.4 — Introduction
29 Apr 2026
AI Safety Atlas Ch.4 — Governance Problems
29 Apr 2026
AI Safety Atlas Ch.4 — Compute Governance
29 Apr 2026
AI Safety Atlas Ch.4 — Systemic Challenges
29 Apr 2026
AI Safety Atlas Ch.4 — Governance Architectures
29 Apr 2026
AI Safety Atlas Ch.4 — Implementation
29 Apr 2026
AI Safety Atlas Ch.4 — Conclusion
29 Apr 2026
AI Safety Atlas Ch.4 — Appendix: Data Governance
29 Apr 2026
AI Safety Atlas Ch.4 — Appendix: National Governance
29 Apr 2026
AI Safety Atlas Ch.5 — Introduction
29 Apr 2026
AI Safety Atlas Ch.5 — Evaluation Design
29 Apr 2026
AI Safety Atlas Ch.5 — Evaluation Techniques
29 Apr 2026
AI Safety Atlas Ch.5 — Evaluated Properties
29 Apr 2026
AI Safety Atlas Ch.5 — Dangerous Capability Evaluations
29 Apr 2026
AI Safety Atlas Ch.5 — Dangerous Propensity Evaluations
29 Apr 2026
AI Safety Atlas Ch.5 — Control Evaluations
29 Apr 2026
AI Safety Atlas Ch.5 — Evaluation Frameworks
29 Apr 2026
AI Safety Atlas Ch.5 — Benchmarks
29 Apr 2026
AI Safety Atlas Ch.5 — Limitations
29 Apr 2026
AI Safety Atlas Ch.5 — Conclusion
29 Apr 2026
AI Safety Atlas Ch.6 — Introduction
29 Apr 2026
AI Safety Atlas Ch.6 — Reinforcement Learning
29 Apr 2026
AI Safety Atlas Ch.6 — Optimization
29 Apr 2026
AI Safety Atlas Ch.6 — Specification Gaming
29 Apr 2026
AI Safety Atlas Ch.6 — Learning from Feedback
29 Apr 2026
AI Safety Atlas Ch.6 — Learning from Imitation
29 Apr 2026
AI Safety Atlas Ch.7 — Introduction
29 Apr 2026
AI Safety Atlas Ch.7 — Learning Dynamics
29 Apr 2026
AI Safety Atlas Ch.7 — Goal-Directedness
29 Apr 2026
AI Safety Atlas Ch.7 — Multi-Objective Generalization
29 Apr 2026
AI Safety Atlas Ch.7 — Scheming
29 Apr 2026
AI Safety Atlas Ch.7 — Detection
29 Apr 2026
AI Safety Atlas Ch.7 — Mitigations
29 Apr 2026
AI Safety Atlas Ch.8 — Introduction
29 Apr 2026
AI Safety Atlas Ch.8 — Oversight
29 Apr 2026
AI Safety Atlas Ch.8 — Task Decomposition
29 Apr 2026
AI Safety Atlas Ch.8 — Process Oversight
29 Apr 2026
AI Safety Atlas Ch.8 — Iterated Amplification
29 Apr 2026
AI Safety Atlas Ch.8 — Debate
29 Apr 2026
AI Safety Atlas Ch.8 — Weak-to-Strong
28 Apr 2026
Alignment Faking in Large Language Models
28 Apr 2026
AI Safety Atlas Ch.1 — Introduction
28 Apr 2026
AI Safety Atlas Ch.1 — Defining and Measuring AGI
28 Apr 2026
AI Safety Atlas Ch.1 — Foundation Models
28 Apr 2026
AI Safety Atlas Ch.1 — Leveraging Scale
28 Apr 2026
AI Safety Atlas Ch.1 — Current Capabilities
28 Apr 2026
AI Safety Atlas Ch.1 — Forecasting Timelines
28 Apr 2026
AI Safety Atlas Ch.1 — Takeoff
28 Apr 2026
AI Safety Atlas Ch.1 — Appendix: Discussion on LLMs
28 Apr 2026
AI Safety Atlas Ch.1 — Appendix: Expert Surveys
28 Apr 2026
AI Safety Atlas Ch.1 — Appendix: Forecasting
28 Apr 2026
AI Safety Atlas Ch.1 — Appendix: Takeoff
28 Apr 2026
AI Safety Atlas Ch.2 — Introduction
28 Apr 2026
AI Safety Atlas Ch.2 — Risk Decomposition
28 Apr 2026
AI Safety Atlas Ch.2 — Dangerous Capabilities
28 Apr 2026
AI Safety Atlas Ch.2 — Risk Amplifiers
28 Apr 2026
AI Safety Atlas Ch.2 — Misuse Risks
28 Apr 2026
AI Safety Atlas Ch.2 — Misalignment Risks
28 Apr 2026
AI Safety Atlas Ch.2 — Systemic Risks
28 Apr 2026
AI Safety Atlas Ch.2 — Conclusion
28 Apr 2026
AI Safety Atlas Ch.2 — Appendix: Forecasting Scenarios
28 Apr 2026
AI Safety Atlas Ch.2 — Appendix: Quantifying Existential Risks
28 Apr 2026
AI Safety Atlas Ch.3 — Introduction
28 Apr 2026
AI Safety Atlas Ch.3 — Definitions
28 Apr 2026
AI Safety Atlas Ch.3 — Challenges
28 Apr 2026
AI Safety Atlas Ch.3 — Misuse Prevention Strategies
28 Apr 2026
AI Safety Atlas Ch.3 — AGI Safety Strategies
28 Apr 2026
AI Safety Atlas Ch.3 — ASI Safety Strategies
28 Apr 2026
AI Safety Atlas Ch.3 — Socio-Technical Strategies
28 Apr 2026
AI Safety Atlas Ch.3 — Combining Strategies
28 Apr 2026
AI Safety Atlas Ch.3 — Conclusion
28 Apr 2026
AI Safety Atlas Ch.3 — Appendix: Long-term Questions
28 Apr 2026
AI Safety Atlas Textbook — Read the Textbook (ai-safety-atlas.com)
28 Apr 2026
Ten AI Safety Projects I'd Like People to Work On — Julian Hazell
27 Apr 2026
100+ concrete projects and open problems in evals
27 Apr 2026
18 Applications of Deception Probes
27 Apr 2026
Incentives for Responsiveness, Instrumental Control and Impact
27 Apr 2026
2404.10636 - What are human values, and how do we align AI to them?
27 Apr 2026
AI Sandbagging: Language Models can Strategically Underperform on Evaluations
27 Apr 2026
What Kind of User Are You? Uncovering User Models in LLM Chatbots
27 Apr 2026
The Perils of Optimizing Learned Reward Functions: Low Training Error Does Not Guarantee Low Regret
27 Apr 2026
WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Models
27 Apr 2026
Future Events as Backdoor Triggers
27 Apr 2026
Tamper-Resistant Safeguards for Open-Weight LLMs
27 Apr 2026
Beyond Preferences in AI Alignment
27 Apr 2026
Robust LLM safeguarding via refusal feature adversarial training
27 Apr 2026
From homeostasis to resource sharing: Biologically and economically aligned multi-objective multi-agent gridworld-based AI safety benchmarks
27 Apr 2026
Challenges and Future Directions of Data-Centric AI Alignment
27 Apr 2026
Gradient Routing: Masking Gradients to Localize Computation in Neural Networks
27 Apr 2026
Rethinking Reward Model Evaluation: Are We Barking up the Wrong Tree?
27 Apr 2026
Improving Steering Vectors by Targeting Sparse Autoencoder Features
27 Apr 2026
On Targeted Manipulation and Deception when Optimizing LLMs for User Feedback
27 Apr 2026
Enhancing Multiple Dimensions of Trustworthiness in LLMs via Sparse Activation Control
27 Apr 2026
Towards evaluations-based safety cases for AI scheming
27 Apr 2026
Rapid Response: Mitigating LLM Jailbreaks with a Few Examples
27 Apr 2026
Do I Know This Entity? Knowledge Awareness and Hallucinations in Language Models
27 Apr 2026
RE-Bench: Evaluating frontier AI R&D capabilities of language model agents against human experts
27 Apr 2026
Adaptive Deployment of Untrusted LLMs Reduces Distributed Threats
27 Apr 2026
The Partially Observable Off-Switch Game
27 Apr 2026
Noise Injection Reveals Hidden Capabilities of Sandbagging Language Models
27 Apr 2026
The Reality of AI and Biorisk
27 Apr 2026
Jailbreak Defense in a Narrow Domain: Limitations of Existing Methods and a New Transcript-Classifier Approach
27 Apr 2026
Frontier Models are Capable of In-context Scheming
27 Apr 2026
Neural Interactive Proofs
27 Apr 2026
Subversion Strategy Eval: Can language models statelessly strategize to subvert control protocols?
27 Apr 2026
Learning Representations of Alignment
27 Apr 2026
Deliberative Alignment: Reasoning Enables Safer Language Models
27 Apr 2026
Observation Interference in Partially Observable Assistance Games
27 Apr 2026
Diverse and Effective Red Teaming with Auto-generated Rewards and Multi-step Reinforcement Learning
27 Apr 2026
Open Problems in Machine Unlearning for AI Safety
27 Apr 2026
Iterative Label Refinement Matters More than Preference Optimization under Weak Supervision
27 Apr 2026
RLHS: Mitigating Misalignment in RLHF with Hindsight Simulation
27 Apr 2026
Infrastructure for AI Agents
27 Apr 2026
Tell me about yourself: LLMs are aware of their learned behaviors
27 Apr 2026
MONA: Myopic Optimization with Non-myopic Approval Can Mitigate Multi-step Reward Hacking
27 Apr 2026
Debate Helps Weak-to-Strong Generalization
27 Apr 2026
Propositional Interpretability in Artificial Intelligence
27 Apr 2026
Open Problems in Mechanistic Interpretability
27 Apr 2026
Gradual Disempowerment
27 Apr 2026
A sketch of an AI control safety case
27 Apr 2026
AI Governance through Markets
27 Apr 2026
Transcoders Beat Sparse Autoencoders for Interpretability
27 Apr 2026
Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming
27 Apr 2026
Defense Against the Dark Prompts: Mitigating Best-of-N Jailbreaking with Prompt Evaluation
27 Apr 2026
AgentBreeder: Mitigating the AI Safety Risks of Multi-Agent Scaffolds via Self-Improvement
27 Apr 2026
Eliciting Language Model Behaviors with Investigator Agents
27 Apr 2026
Robust LLM Alignment via Distributionally Robust Direct Preference Optimization
27 Apr 2026
The Elicitation Game: Evaluating Capability Elicitation Techniques
27 Apr 2026
Detecting Strategic Deception Using Linear Probes
27 Apr 2026
Toward universal steering and monitoring of AI models
27 Apr 2026
Great Models Think Alike and this Undermines AI Oversight
27 Apr 2026
Scalable Oversight for Superhuman AI via Recursive Self-Critiquing
27 Apr 2026
Model Tampering Attacks Enable More Rigorous Evaluations of LLM Capabilities
27 Apr 2026
You Are What You Eat -- AI Alignment Requires Understanding How Data Shapes Structure and Generalisation
27 Apr 2026
Assessing confidence in frontier AI safety cases
27 Apr 2026
Compromising Honesty and Harmlessness in Language Models via Deception Attacks
27 Apr 2026
Utility Engineering: Analyzing and Controlling Emergent Value Systems in AIs
27 Apr 2026
AI Alignment at Your Discretion
27 Apr 2026
Adversarial Alignment for LLMs Requires Simpler, Reproducible, and More Measurable Objectives
27 Apr 2026
Demonstrating specification gaming in reasoning models
27 Apr 2026
Safe Learning Under Irreversible Dynamics via Asking for Help
27 Apr 2026
Multi-Agent Risks from Advanced AI
27 Apr 2026
Fundamental Limitations in Pointwise Defences of LLM Finetuning APIs
27 Apr 2026
Superintelligent Agents Pose Catastrophic Risks: Can Scientist AI Offer a Safer Path?
27 Apr 2026
Forecasting Frontier Language Model Agent Capabilities
27 Apr 2026
Forecasting Rare Language Model Behaviors
27 Apr 2026
REINFORCE Adversarial Attacks on Large Language Models: An Adaptive, Distributional, and Semantic Objective
27 Apr 2026
The Geometry of Refusal in Large Language Models: Concept Cones and Representational Independence
27 Apr 2026
Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs
27 Apr 2026
No, of Course I Can! Deeper Fine-Tuning Attacks That Bypass Token-Level Safety Mechanisms
27 Apr 2026
Modeling Human Beliefs about AI Behavior for Scalable Oversight
27 Apr 2026
Societal alignment frameworks can improve llm alignment
27 Apr 2026
Steering Large Language Model Activations in Sparse Spaces
27 Apr 2026
Teaching AI to Handle Exceptions: Supervised Fine-tuning with Human-aligned Judgment
27 Apr 2026
Maximizing Signal in Human-Model Preference Alignment
27 Apr 2026
AILuminate: Introducing v1.0 of the AI Risk and Reliability Benchmark from MLCommons
27 Apr 2026
SAEBench: A Comprehensive Benchmark for Sparse Autoencoders in Language Model Interpretability
27 Apr 2026
Auditing language models for hidden objectives
27 Apr 2026
Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation
27 Apr 2026
Superalignment with Dynamic Human Values
27 Apr 2026
MMDT: Decoding the Trustworthiness and Safety of Multimodal Foundation Models
27 Apr 2026
Large language model-powered AI systems achieve self-replication with no human intervention
27 Apr 2026
An Approach to Technical AGI Safety and Security
27 Apr 2026
Interpreting Emergent Planning in Model-Free Reinforcement Learning
27 Apr 2026
Overcoming Sparsity Artifacts in Crosscoders to Interpret Chat-Tuning
27 Apr 2026
Among Us: A Sandbox for Measuring and Detecting Agentic Deception
27 Apr 2026
How to evaluate control measures for LLM agents? A trajectory from today to superintelligence
27 Apr 2026
AssistanceZero: Scalably Solving Assistance Games
27 Apr 2026
Mechanistic Anomaly Detection for "Quirky" Language Models
27 Apr 2026
Mitigating Many-Shot Jailbreaking
27 Apr 2026
The Structural Safety Generalization Problem
27 Apr 2026
Evaluating the Goal-Directedness of Large Language Models
27 Apr 2026
X-Teaming: Multi-Turn Jailbreaks and Defenses with Adaptive Multi-Agents
27 Apr 2026
The Geometry of Self-Verification in a Task-Specific Reasoning Model
27 Apr 2026
Virology Capabilities Test (VCT): A Multimodal Virology Q&A Benchmark
27 Apr 2026
Safety Pretraining: Toward the Next Generation of Safe AI
27 Apr 2026
Scaling Laws For Scalable Oversight
27 Apr 2026
AI Awareness (literature review)
27 Apr 2026
Evaluating Frontier Models for Stealth and Situational Awareness
27 Apr 2026
Technical Report: Evaluating Goal Drift in Language Model Agents
27 Apr 2026
An alignment safety case sketch based on debate
27 Apr 2026
Societal and technological progress as sewing an ever-growing, ever-changing, patchy, and polychrome quilt
27 Apr 2026
Reasoning Models Don't Always Say What They Think
27 Apr 2026
Safety by Measurement: A Systematic Literature Review of AI Safety Evaluation Methods
27 Apr 2026
LLMs Outperform Experts on Challenging Biology Benchmarks
27 Apr 2026
Winning at All Cost: A Small Environment for Eliciting Specification Gaming Behaviors in Large Language Models
27 Apr 2026
Preference Learning with Lie Detectors can Induce Honesty or Evasion
27 Apr 2026
Towards eliciting latent knowledge from LLMs with mechanistic interpretability
27 Apr 2026
Will AI Tell Lies to Save Sick Children? Litmus-Testing AI Values Prioritization with AIRiskDilemmas
27 Apr 2026
Unlearning Isn't Deletion: Investigating Reversibility of Machine Unlearning in LLMs
27 Apr 2026
Discovering Forbidden Topics in Language Models
27 Apr 2026
Safety Alignment via Constrained Knowledge Unlearning
27 Apr 2026
When Ethics and Payoffs Diverge: LLM Agents in Morally Charged Social Dilemmas
27 Apr 2026
Shutdownable Agents through POST-Agency
27 Apr 2026
Deep Reinforcement Learning Agents are not even close to Human Intelligence
27 Apr 2026
CoT Red-Handed: Stress Testing Chain-of-Thought Monitoring
27 Apr 2026
Large Language Models Often Know When They Are Being Evaluated
27 Apr 2026
Large language models can learn and generalize steganographic chain-of-thought under process supervision
27 Apr 2026
AI Debate Aids Assessment of Controversial Claims
27 Apr 2026
It's the Thought that Counts: Evaluating the Attempts of Frontier LLMs to Persuade on Harmful Topics
27 Apr 2026
When Thinking LLMs Lie: Unveiling the Strategic Deception in Representations of Reasoning Models
27 Apr 2026
Preference Learning for AI Alignment: a Causal Perspective
27 Apr 2026
Distillation Robustifies Unlearning
27 Apr 2026
Reward Model Interpretability via Optimal and Pessimal Tokens
27 Apr 2026
Detecting High-Stakes Interactions with Activation Probes
27 Apr 2026
Decomposing MLP Activations into Interpretable Features via Semi-Nonnegative Matrix Factorization
27 Apr 2026
Monitoring Decomposition Attacks in LLMs with Lightweight Sequential Monitors
27 Apr 2026
RedDebate: Safer Responses through Multi-Agent Red Teaming Debates
27 Apr 2026
Model Organisms for Emergent Misalignment
27 Apr 2026
Convergent Linear Representations of Emergent Misalignment
27 Apr 2026
Robust LLM Unlearning with MUDMAN: Meta-Unlearning with Disruption Masking And Normalization
27 Apr 2026
Jailbreak Transferability Emerges from Shared Representations
27 Apr 2026
Thought Crime: Backdoors and Emergent Misalignment in Reasoning Models
27 Apr 2026
RL-Obfuscation: Can Language Models Learn to Evade Latent-Space Monitors?
27 Apr 2026
OS-Harm: A Benchmark for Measuring Safety of Computer Use Agents
27 Apr 2026
SHADE-Arena: Evaluating Sabotage and Monitoring in LLM Agents
27 Apr 2026
2506.17434 - Resource Rational Contractualism Should Guide AI Alignment
27 Apr 2026
Why Do Some Language Models Fake Alignment While Others Don't?
27 Apr 2026
Why Do Some Language Models Fake Alignment While Others Don't?
27 Apr 2026
Inference-Time Reward Hacking in Large Language Models
27 Apr 2026
Persona Features Control Emergent Misalignment
27 Apr 2026
A Different Approach to AI Safety Proceedings from the Columbia Convening on AI Openness and Safety
27 Apr 2026
Report on NSF Workshop on Science of Safe AI
27 Apr 2026
Teaching Models to Verbalize Reward Hacking in Chain-of-Thought Reasoning
27 Apr 2026
STACK: Adversarial Attacks on LLM Safeguard Pipelines
27 Apr 2026
Early Signs of Steganographic Capabilities in Frontier LLMs
27 Apr 2026
'For Argument's Sake, Show Me How to Harm Myself!': Jailbreaking LLMs in Suicide and Self-Harm Contexts
27 Apr 2026
Mitigating Goal Misgeneralization via Minimax Regret
27 Apr 2026
Adversarial Manipulation of Reasoning Models using Internal Representations
27 Apr 2026
Lessons from a Chimp: AI "Scheming" and the Quest for Ape Language
27 Apr 2026
When Chain of Thought is Necessary, Language Models Struggle to Evade Monitors
27 Apr 2026
OpenAgentSafety: A Comprehensive Framework for Evaluating Real-World AI Agent Safety
27 Apr 2026
Misalignment From Treating Means as Ends
27 Apr 2026
Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety
27 Apr 2026
The Safety Gap Toolkit: Evaluating Hidden Dangers of Open-Source Models
27 Apr 2026
Jailbreak-Tuning: Models Efficiently Learn Jailbreak Susceptibility
27 Apr 2026
Manipulation Attacks by Misaligned AI: Risk Analysis and Safety Case Framework
27 Apr 2026
When Autonomy Goes Rogue: Preparing for Risks of Multi-Agent Collusion in Social Systems
27 Apr 2026
Subliminal Learning: Language models transmit behavioural traits via hidden signals in data
27 Apr 2026
Steering Out-of-Distribution Generalization with Concept Ablation Fine-Tuning
27 Apr 2026
Persona Vectors: Monitoring and Controlling Character Traits in Language Models
27 Apr 2026
When Truthful Representations Flip Under Deceptive Instructions?
27 Apr 2026
Model-Based Soft Maximization of Suitable Metrics of Long-Term Human Power
27 Apr 2026
LLMs Can Covertly Sandbag on Capability Evaluations Against Chain-of-Thought Monitoring
27 Apr 2026
LLM Robustness Leaderboard v1 --Technical report
27 Apr 2026
Deep Ignorance: Filtering Pretraining Data Builds Tamper-Resistant Safeguards into Open-Weight LLMs
27 Apr 2026
The PacifAIst Benchmark:Would an Artificial Intelligence Choose to Sacrifice Itself for Human Safety?
27 Apr 2026
Rethinking Safety in LLM Fine-tuning: An Optimization Perspective
27 Apr 2026
CRISP: Persistent Concept Unlearning via Sparse Autoencoders
27 Apr 2026
AI Testing Should Account for Sophisticated Strategic Behaviour
27 Apr 2026
Limit-Computable Grains of Truth for Arbitrary Computable Extensive-Form (Un)Known Games
27 Apr 2026
School of Reward Hacks: Hacking harmless tasks generalizes to misaligned behavior in LLMs
27 Apr 2026
Caught in the Act: a mechanistic approach to detecting deception
27 Apr 2026
Evaluating Language Model Reasoning about Confidential Information
27 Apr 2026
Probe-Rewrite-Evaluate: A Workflow for Reliable Benchmarks and Quantifying Evaluation Awareness
27 Apr 2026
EigenBench: A Comparative behavioural Measure of Value Alignment
27 Apr 2026
BioBlue: Notable runaway-optimiser-like LLM failure modes on biologically and economically aligned AI safety benchmarks for LLMs with simplified observation format
27 Apr 2026
Murphys Laws of AI Alignment: Why the Gap Always Wins
27 Apr 2026
Virtual Agent Economies
27 Apr 2026
Collapse of Irrelevant Representations (CIR) Ensures Robust and Non-Disruptive LLM Unlearning
27 Apr 2026
Shutdown Resistance in Large Language Models
27 Apr 2026
Stress Testing Deliberative Alignment for Anti-Scheming Training
27 Apr 2026
Stress Testing Deliberative Alignment for Anti-Scheming Training
27 Apr 2026
D-REX: A Benchmark for Detecting Deceptive Reasoning in Large Language Models
27 Apr 2026
Strategic Dishonesty Can Undermine AI Safety Evaluations of Frontier LLMs
27 Apr 2026
Active Attacks: Red-teaming LLMs via Adaptive Environments
27 Apr 2026
Multiplayer Nash Preference Optimization
27 Apr 2026
A Realistic Evaluation of Self-Replication Risk in LLM Agents
27 Apr 2026
Generative Value Conflicts Reveal LLM Priorities
27 Apr 2026
Binary Sparse Coding for Interpretability
27 Apr 2026
Beyond Linear Probes: Dynamic Safety Monitoring for Language Models
27 Apr 2026
Sandbagging in a Simple Survival Bandit Problem
27 Apr 2026
Large Reasoning Models Learn Better Alignment from Flawed Thinking
27 Apr 2026
Eliciting Secret Knowledge from Language Models
27 Apr 2026
Is It Thinking or Cheating? Detecting Implicit Reward Hacking by Measuring Reasoning Effort
27 Apr 2026
InvThink: Towards AI Safety via Inverse Reasoning
27 Apr 2026
Where Did It Go Wrong? Attributing Undesirable LLM Behaviors via Representation Gradient Tracing
27 Apr 2026
Inoculation Prompting: Eliciting traits from LLMs during training can suppress them at test-time
27 Apr 2026
Inoculation Prompting: Instructing LLMs to misbehave at train-time improves test-time alignment
27 Apr 2026
Agentic Misalignment
27 Apr 2026
Moloch's Bargain: Emergent Misalignment When LLMs Compete for Audiences
27 Apr 2026
Stress-Testing Model Specs Reveals Character Differences among Language Models
27 Apr 2026
The Unintended Trade-off of AI Alignment:Balancing Hallucination Mitigation and Safety in LLMs
27 Apr 2026
The Alignment Waltz: Jointly Training Agents to Collaborate for Safety
27 Apr 2026
Keep Calm and Avoid Harmful Content: Concept Alignment and Latent Manipulation Towards Safer Answers
27 Apr 2026
Scheming Ability in LLM-to-LLM Strategic Interactions
27 Apr 2026
Training LLM Agents to Empower Humans
27 Apr 2026
Narrow Finetuning Leaves Clearly Readable Traces in Activation Differences
27 Apr 2026
Evaluating & Reducing Deceptive Dialogue From Language Models with Multi-turn RL
27 Apr 2026
DECEPTIONBENCH: A Comprehensive Benchmark for AI Deception Behaviors in Real-world Scenario
27 Apr 2026
A Definition of AGI
27 Apr 2026
Can Reasoning Models Obfuscate Reasoning? Stress-Testing Chain-of-Thought Monitorability
27 Apr 2026
LLMs can hide text in other text of the same length
27 Apr 2026
Steering Evaluation-Aware Language Models to Act Like They Are Deployed
27 Apr 2026
Reducing the Probability of Undesirable Outputs in Language Models Using Probabilistic Inference
27 Apr 2026
Toward Understanding the Transferability of Adversarial Suffixes in Large Language Models
27 Apr 2026
A Pragmatic Way to Measure Chain-of-Thought Monitorability
27 Apr 2026
What's In My Human Feedback? Learning Interpretable Descriptions of Preference Data
27 Apr 2026
Chain-of-Thought Hijacking
27 Apr 2026
Consistency Training Helps Stop Sycophancy and Jailbreaks
27 Apr 2026
Best Practices for Biorisk Evaluations on Open-Weight Bio-Foundation Models
27 Apr 2026
Best Practices for Biorisk Evaluations on Open-Weight Bio-Foundation Models
27 Apr 2026
Open Character Training: Shaping the Persona of AI Assistants through Constitutional AI
27 Apr 2026
Optimizing AI Agent Attacks With Synthetic Data
27 Apr 2026
Evaluating Control Protocols for Untrusted AI Agents
27 Apr 2026
Steering Language Models with Weight Arithmetic
27 Apr 2026
Spilling the Beans: Teaching LLMs to Self-Report Their Hidden Objectives
27 Apr 2026
Rank-1 LoRAs Encode Interpretable Reasoning Signals
27 Apr 2026
Training Language Models to Explain Their Own Computations
27 Apr 2026
Liars' Bench: Evaluating Lie Detectors for Language Models
27 Apr 2026
Liars' Bench: Evaluating Lie Detectors for Language Models
27 Apr 2026
Difficulties with Evaluating a Deception Detector for AIs
27 Apr 2026
Symmetries at the origin of hierarchical emergence
27 Apr 2026
In-Context Representation Hijacking
27 Apr 2026
Beyond Data Filtering: Knowledge Localization for Capability Removal in LLMs
27 Apr 2026
Training LLMs for Honesty via Confessions
27 Apr 2026
A computational no-coincidence principle
27 Apr 2026
A Concrete Roadmap towards Safety Cases based on Chain-of-Thought Monitoring
27 Apr 2026
A dataset of rated conceptual arguments
27 Apr 2026
A Pragmatic Vision for Interpretability
27 Apr 2026
A Pragmatic Vision for Interpretability
27 Apr 2026
A Safety Case for a Deployed LLM: Corrigibility as a Singular Target
27 Apr 2026
A Shutdown Problem Proposal
27 Apr 2026
A single principle related to many Alignment subproblems?
27 Apr 2026
A Three-Layer Model of LLM Psychology
27 Apr 2026
A Toy Evaluation of Inference Code Tampering
27 Apr 2026
Abstract Mathematical Concepts vs. Abstractions Over Real-World Systems
27 Apr 2026
ADeLe v1.0: A battery for AI Evaluation with explanatory and predictive power
27 Apr 2026
Advancing Gemini's security safeguards
27 Apr 2026
Aesthetic Preferences Can Cause Emergent Misalignment
27 Apr 2026
Aether July 2025 Update
27 Apr 2026
Against blanket arguments against interpretability
27 Apr 2026
Against RL: The Case for System 2 Learning
27 Apr 2026
Agent foundations: not really math, not really science
27 Apr 2026
Agentic Interpretability: A Strategy Against Gradual Disempowerment
27 Apr 2026
Agentic Misalignment: How LLMs Could be Insider Threats
27 Apr 2026
Agentic Misalignment: How LLMs could be insider threats
27 Apr 2026
Agentic Misalignment: How LLMs could be insider threats
27 Apr 2026
AI Assistants Should Have a Direct Line to Their Developers
27 Apr 2026
AI companies are unlikely to make high-assurance safety cases if timelines are short
27 Apr 2026
AI companies should be safety-testing the most capable versions of their models
27 Apr 2026
Summary: AI Safety (Wikipedia)
27 Apr 2026
AIs at the current capability level may be important for future safety work
27 Apr 2026
Aligning machine and human visual representations across abstraction levels
27 Apr 2026
Alignment Faking Revisited: Improved Classifiers and Open Source Extensions
27 Apr 2026
Alignment first, intelligence later
27 Apr 2026
Ambiguous Online Learning
27 Apr 2026
An alignment safety case sketch based on debate
27 Apr 2026
Auditing language models for hidden objectives
27 Apr 2026
Automated Researchers Can Subtly Sandbag
27 Apr 2026
Automatically Jailbreaking Frontier Language Models with Investigator Agents
27 Apr 2026
Automating AI Safety: What we can do today
27 Apr 2026
Bare Minimum Mitigations for Autonomous AI Development
27 Apr 2026
Beliefs about formal methods and AI safety
27 Apr 2026
Benchmarking deception probes for trusted monitoring
27 Apr 2026
Bridging the human–AI knowledge gap through concept discovery and transfer in AlphaZero
27 Apr 2026
Building and evaluating alignment auditing agents
27 Apr 2026
Building and evaluating alignment auditing agents
27 Apr 2026
Call for Collaboration: Renormalization for AI safety
27 Apr 2026
Call Me A Jerk: Persuading AI to Comply with Objectionable Requests
27 Apr 2026
Can a Neural Network that only Memorizes the Dataset be Undetectably Backdoored?
27 Apr 2026
CCS-Lib: A Python package to elicit latent knowledge from LLMs
27 Apr 2026
Chain-of-Thought Snippets — Anti-Scheming
27 Apr 2026
Circuit Tracing: Revealing Computational Graphs in Language Models
27 Apr 2026
Circuits in Superposition 2: Now With Less Wrong Math
27 Apr 2026
Circuits in Superposition
27 Apr 2026
Clarifying "wisdom": Foundational topics for aligned AIs to prioritize before irreversible decisions
27 Apr 2026
Claude's Constitution
27 Apr 2026
Claude Sonnet 3.7 (often) knows when it's in alignment evaluations
27 Apr 2026
Code World Model Preparedness Report
27 Apr 2026
Collective cooperative intelligence
27 Apr 2026
Communication & Trust
27 Apr 2026
Competing with sampling
27 Apr 2026
Compressed Computation is (probably) not Computation in Superposition
27 Apr 2026
Condensation
27 Apr 2026
Constitutional Classifiers: Defending against universal jailbreaks
27 Apr 2026
Control protocols don't always need to know which models are scheming
27 Apr 2026
ControlArena
27 Apr 2026
Convergent Linear Representations of Emergent Misalignment
27 Apr 2026
Cost-Effective Constitutional Classifiers via Representation Re-use
27 Apr 2026
CoT May Be Highly Informative Despite "Unfaithfulness"
27 Apr 2026
Cross-Architecture Model Diffing with Crosscoders: Unsupervised Discovery of Differences Between LLMs
27 Apr 2026
Ctrl-Z: Controlling AI Agents via Resampling
27 Apr 2026
CyberSecEval 4
27 Apr 2026
Deceptive Alignment and Homuncularity
27 Apr 2026
Deep ignorance: Filtering pretraining data builds tamper-resistant safeguards into open-weight LLMs
27 Apr 2026
Details about METR's evaluation of OpenAI GPT-5
27 Apr 2026
Details about METR's preliminary evaluation of OpenAI's o3 and o4-mini
27 Apr 2026
Detect Goodhart and shut down
27 Apr 2026
Detecting and reducing scheming in AI models
27 Apr 2026
Detecting misbehavior in frontier reasoning models
27 Apr 2026
Detecting Strategic Deception Using Linear Probes
27 Apr 2026
Discovering Undesired Rare Behaviors via Model Diff Amplification
27 Apr 2026
Discovering Undesired Rare Behaviors via Model Diff Amplification
27 Apr 2026
Do LLMs Comply Differently During Tests? Is This a Hidden Variable in Safety Evaluation? And Can We Steer That?
27 Apr 2026
Do LLMs know what they're capable of? Why this matters for AI safety, and initial findings
27 Apr 2026
Do Not Tile the Lightcone with Your Confused Ontology
27 Apr 2026
Do reasoning models use their scratchpad like we do? Evidence from distilling paraphrases
27 Apr 2026
Do safety-relevant LLM steering vectors optimized on a single example generalize?
27 Apr 2026
Docent: A system for analyzing and intervening on agent behavior
27 Apr 2026
Dodging systematic human errors in scalable oversight
27 Apr 2026
Dodging systematic human errors in scalable oversight
27 Apr 2026
Edge Cases in AI Alignment
27 Apr 2026
Emergent Misalignment on a Budget
27 Apr 2026
Emergent Misalignment & Realignment
27 Apr 2026
Emmett Shear on Building AI That Actually Cares: Beyond Control and Steering
27 Apr 2026
End A Subset Of Conversations
27 Apr 2026
Enhancing Model Safety through Pretraining Data Filtering
27 Apr 2026
Evaluating honesty and lie detection techniques on a diverse suite of dishonest models
27 Apr 2026
Evaluating potential cybersecurity threats of advanced AI
27 Apr 2026
Findings from a Pilot Anthropic—OpenAI Alignment Evaluation Exercise
27 Apr 2026
Findings from a Pilot Anthropic—OpenAI Alignment Evaluation Exercise
27 Apr 2026
Findings from a pilot Anthropic–OpenAI alignment evaluation exercise: OpenAI Safety Tests
27 Apr 2026
Formalizing Embeddedness Failures in Universal Artificial Intelligence
27 Apr 2026
From SLT to AIT: NN Generalisation Out of Distribution
27 Apr 2026
Full-Stack Alignment
27 Apr 2026
Games for AI Control
27 Apr 2026
Giving AIs safe motivations
27 Apr 2026
Here's 18 Applications of Deception Probes
27 Apr 2026
Here's 18 Applications of Deception Probes
27 Apr 2026
HIBP Human Inductive Bias Project Plan
27 Apr 2026
Hierarchical Alignment
27 Apr 2026
High Actuation Spaces - Sahil
27 Apr 2026
How Can Interpretability Researchers Help AGI Go Well?
27 Apr 2026
How Does Time Horizon Vary Across Domains?
27 Apr 2026
How important is the model spec if alignment fails?
27 Apr 2026
I replicated the Anthropic alignment faking experiment on other models, and they didn't fake alignment
27 Apr 2026
Illusory Safety: Redteaming DeepSeek R1 and the Strongest Fine-Tunable Models of OpenAI, Anthropic, and Google
27 Apr 2026
Imitation learning is probably existentially safe
27 Apr 2026
Infra-Bayesian Decision-Estimation Theory
27 Apr 2026
Infra-Bayesianism category on LessWrong
27 Apr 2026
Insights on Crosscoder Model Diffing
27 Apr 2026
Inspect Cyber
27 Apr 2026
Inspect Evals
27 Apr 2026
Instrumental Goals Are A Different And Friendlier Kind Of Thing Than Terminal Goals
27 Apr 2026
Introducing Anthropic's Safeguards Research Team
27 Apr 2026
Investigating truthfulness in a pre-release o3 model
27 Apr 2026
Is alignment reducible to becoming more coherent?
27 Apr 2026
It's hard to make scheming evals look realistic for LLMs
27 Apr 2026
Live Theory
27 Apr 2026
LLM AGI may reason about its goals and discover misalignments by default
27 Apr 2026
LLM AGI will have memory, and memory changes alignment
27 Apr 2026
Luthien's Approach to Prosaic AI Control in 21 Points
27 Apr 2026
Maintaining Alignment during RSI as a Feedback Control Problem
27 Apr 2026
MALT: A Dataset of Natural and Prompted Behaviors That Threaten Eval Integrity
27 Apr 2026
Measuring AI Ability to Complete Long Tasks
27 Apr 2026
Misalignment and Strategic Underperformance: An Analysis of Sandbagging and Exploration Hacking
27 Apr 2026
Mistral Large 2 (123B) seems to exhibit alignment faking
27 Apr 2026
Model Integrity
27 Apr 2026
Modifying LLM Beliefs with Synthetic Document Finetuning
27 Apr 2026
MONA: Managed Myopia with Approval Feedback
27 Apr 2026
Monitoring computer use via hierarchical summarization
27 Apr 2026
MoSSAIC: AI Safety After Mechanism
27 Apr 2026
Multipolar AI is Underrated
27 Apr 2026
Narrow Misalignment is Hard, Emergent Misalignment is Easy
27 Apr 2026
Negative Results for SAEs On Downstream Tasks and Deprioritising SAE Research (GDM Mech Interp Team Progress Update #2)
27 Apr 2026
Neural Interactive Proofs
27 Apr 2026
New website analyzing AI companies' model evals
27 Apr 2026
No-self as an alignment target
27 Apr 2026
Non-Monotonic Infra-Bayesian Physicalism
27 Apr 2026
Obstacles in ARC's research agenda
27 Apr 2026
Off-switching not guaranteed
27 Apr 2026
On Eudaimonia and Optimization
27 Apr 2026
On the Biology of a Large Language Model
27 Apr 2026
On the functional self of LLMs
27 Apr 2026
One-shot steering vectors cause emergent misalignment, too
27 Apr 2026
Open Philanthropy Technical AI Safety RFP - $40M Available Across 21 Research Areas
27 Apr 2026
Open problems in emergent misalignment
27 Apr 2026
Open Source Replication of Anthropic's Crosscoder paper for model-diffing
27 Apr 2026
Open-sourcing circuit tracing tools
27 Apr 2026
Open Technical Problems in Open-Weight AI Model Risk Management
27 Apr 2026
OpenAgentSafety: A Comprehensive Framework for Evaluating Real-World AI Agent Safety
27 Apr 2026
OpenAI Model Spec
27 Apr 2026
Opportunity Space: Renormalization for AI Safety
27 Apr 2026
Opus 4.5's Soul Document
27 Apr 2026
Our updated Preparedness Framework
27 Apr 2026
PaperBench: Evaluating AI's Ability to Replicate AI Research
27 Apr 2026
Perils of Under vs Over-sculpting AGI Desires
27 Apr 2026
Petri: An open-source auditing tool to accelerate AI safety research
27 Apr 2026
Petri: An open-source auditing tool to accelerate AI safety research
27 Apr 2026
Petri: An open-source auditing tool to accelerate AI safety research
27 Apr 2026
Petri: An open-source auditing tool to accelerate AI safety research
27 Apr 2026
Platonic representation hypothesis
27 Apr 2026
Preference gaps as a safeguard against AI self-replication
27 Apr 2026
Principles for Picking Practical Interpretability Projects
27 Apr 2026
Problems with instruction-following as an alignment target
27 Apr 2026
Prospects for Alignment Automation: Interpretability Case Study
27 Apr 2026
Prover-Estimator Debate: A New Scalable Oversight Protocol
27 Apr 2026
Psychopathia Machinalis: A Nosological Framework for Understanding Pathologies in Advanced Artificial Intelligence
27 Apr 2026
Putting up Bumpers
27 Apr 2026
Realistic Reward Hacking Induces Different and Deeper Misalignment
27 Apr 2026
Reasoning models don't always say what they think
27 Apr 2026
Recommendations for Technical AI Safety Research Directions
27 Apr 2026
Recontextualization Mitigates Specification Gaming Without Modifying the Specification
27 Apr 2026
Reducing LLM deception at scale with self-other overlap fine-tuning
27 Apr 2026
Reimagining Alignment
27 Apr 2026
RepliBench: measuring autonomous replication capabilities in AI systems
27 Apr 2026
Report & retrospective on the Dovetail fellowship
27 Apr 2026
Research Agenda for Sociotechnical Approaches to AI Safety
27 Apr 2026
Research directions Open Phil wants to fund in technical AI safety
27 Apr 2026
Research Note: Our scheming precursor evals had limited predictive power for our in-context scheming evals
27 Apr 2026
Research Note: Our scheming precursor evals had limited predictive power for our in-context scheming evals
27 Apr 2026
Reward button alignment
27 Apr 2026
Rosas
27 Apr 2026
Safe (Pareto) Improvements in Binary Constraint Structures
27 Apr 2026
SafePlanBench: evaluating a Guaranteed Safe AI Approach for LLM-based Agents
27 Apr 2026
Safety by Measurement: A Systematic Literature Review of AI Safety Evaluation Methods
27 Apr 2026
Safety cases for Pessimism
27 Apr 2026
Safety evaluations hub
27 Apr 2026
Scaling Sparse Feature Circuit Finding to Gemma 9B
27 Apr 2026
Selection Pressures on LM Personas
27 Apr 2026
Selective Generalization: Improving Capabilities While Maintaining Alignment
27 Apr 2026
Selective modularity: a research agenda
27 Apr 2026
Self-Fulfilling Misalignment Data Might Be Poisoning Our AI Models
27 Apr 2026
Self-preservation or Instruction Ambiguity? Examining the Causes of Shutdown Resistance
27 Apr 2026
Serious Flaws in CAST
27 Apr 2026
SHADE-Arena: Evaluating sabotage and monitoring in LLM agents
27 Apr 2026
Shutdownable Agents through POST-Agency
27 Apr 2026
Six Thoughts on AI Safety
27 Apr 2026
SLT for AI Safety
27 Apr 2026
Spiral-Bench
27 Apr 2026
Subliminal Learning: Language Models Transmit behavioural Traits via Hidden Signals in Data
27 Apr 2026
Surfacing Pathological Behaviors in Language Models
27 Apr 2026
System 2 Alignment: Deliberation, Review, and Thought Management
27 Apr 2026
Systematic runaway-optimiser-like LLM failure modes on Biologically and Economically aligned AI safety benchmarks for LLMs with simplified observation format (BioBlue)
27 Apr 2026
Takeaways from sketching a control safety case
27 Apr 2026
Taking a responsible path to AGI
27 Apr 2026
Testing for Scheming with Model Deletion
27 Apr 2026
The Alignment Project by UK AISI
27 Apr 2026
The Circuits Research Landscape
27 Apr 2026
"The Era of Experience" has an unsolved technical alignment problem
27 Apr 2026
The Frame-Dependent Mind
27 Apr 2026
The MASK Evaluation
27 Apr 2026
The Pando Problem: Rethinking AI Individuality
27 Apr 2026
The Rise of Parasitic AI
27 Apr 2026
The Rise of Parasitic AI
27 Apr 2026
The Soul Document
27 Apr 2026
The Urgency of Interpretability
27 Apr 2026
the void
27 Apr 2026
Three Sketches of ASL-4 Safety Case Components
27 Apr 2026
Toward understanding and preventing misalignment generalization
27 Apr 2026
Towards a Scale-Free Theory of Intelligent Agency
27 Apr 2026
Towards Alignment Auditing as a Numbers-Go-Up Science
27 Apr 2026
Trading Inference-Time Compute for Adversarial Robustness
27 Apr 2026
Training AI to do alignment research we don't already know how to do
27 Apr 2026
Training fails to elicit subtle reasoning in current language models
27 Apr 2026
Training on Documents About Reward Hacking Induces Reward Hacking
27 Apr 2026
UAIASI
27 Apr 2026
UK AISI Alignment Team: Debate Sequence
27 Apr 2026
UK AISI's Alignment Team: Research Agenda
27 Apr 2026
Understanding and Controlling LLM Generalization
27 Apr 2026
Understanding the Capabilities and Limitations of Weak-to-Strong Generalization
27 Apr 2026
Unlearning Needs to be More Selective [Progress Report]
27 Apr 2026
Unsupervised Elicitation
27 Apr 2026
Validating against a misalignment detector is very different to training against one
27 Apr 2026
Video and transcript of talk on automating alignment research
27 Apr 2026
We need a field of Reward Function Design
27 Apr 2026
We should try to automate AI safety work asap
27 Apr 2026
Weak to Strong Generalization for Large Language Models with Multi-capabilities
27 Apr 2026
What, if not agency?
27 Apr 2026
What is Inadequate about Bayesianism for AI Alignment: Motivating Infra-Bayesianism
27 Apr 2026
What Is The Alignment Problem?
27 Apr 2026
What We Learned Trying to Diff Base and Chat Models (And Why It Matters)
27 Apr 2026
When does Claude sabotage code? An Agentic Misalignment follow-up
27 Apr 2026
White Box Control at UK AISI - Update on Sandbagging Investigations
27 Apr 2026
Whitebox detection of sandbagging model organisms
27 Apr 2026
Why Corrigibility is Hard and Important (i.e. "Whence the high MIRI confidence in alignment difficulty?")
27 Apr 2026
Why do misalignment risks increase as AIs get more capable?
27 Apr 2026
Why Do Some Language Models Fake Alignment While Others Don't?
27 Apr 2026
Why Future AIs will Require New Alignment Methods
27 Apr 2026
Why it's good for AI reasoning to be legible and faithful
27 Apr 2026
Why modelling multi-objective homeostasis is essential for AI alignment (and how it helps with AI safety as well). Subtleties and Open Challenges
27 Apr 2026
Will alignment-faking Claude accept a deal to reveal its misalignment?
27 Apr 2026
Won't vs. Can't: Sandbagging-like Behavior from Claude Models
27 Apr 2026
Working through a small tiling result
23 Apr 2026
Examining Popular Arguments Against AI Existential Risk: A Philosophical Analysis
16 Apr 2026
How to Prepare for AGI — Benjamin Todd
08 Apr 2026
Summary: 80,000 Hours — Why AI Risks Are the World's Most Pressing Problems
08 Apr 2026
Summary: 80,000 Hours Career Guide — What Makes for a Dream Job?
08 Apr 2026
Summary: 80,000 Hours Career Guide — Overview & Introduction
08 Apr 2026
Summary: 80,000 Hours — Catastrophic AI Misuse
08 Apr 2026
Summary: 80,000 Hours — Extreme Power Concentration
08 Apr 2026
Summary: 80,000 Hours — Why Problem Choice Is the Biggest Driver of Impact
08 Apr 2026
Summary: 80,000 Hours Podcast — Ajeya Cotra on Accidentally Teaching AI to Deceive Us
08 Apr 2026
Summary: 80,000 Hours Podcast — Ajeya Cotra on Transformative AI Crunch Time
08 Apr 2026
Summary: 80,000 Hours Podcast — Ben Garfinkel on Scrutinising Classic AI Risk Arguments
08 Apr 2026
Summary: 80,000 Hours Podcast — Buck Shlegeris on AI Control and Scheming
08 Apr 2026
Summary: 80,000 Hours Podcast — Holden Karnofsky on How AI Could Take Over the World
08 Apr 2026
Summary: 80,000 Hours Podcast — Holden Karnofsky on Concrete AI Safety at Frontier Companies
08 Apr 2026
Summary: 80,000 Hours Podcast — AI Safety Episode Index
08 Apr 2026
Summary: 80,000 Hours Podcast — Jan Leike on Superalignment
08 Apr 2026
Summary: 80,000 Hours Podcast — Nick Joseph on Anthropic's Safety Approach
08 Apr 2026
Summary: 80,000 Hours Podcast — Nova DasSarma on Information Security and AI
08 Apr 2026
Summary: 80,000 Hours Podcast — Catherine Olsson & Daniel Ziegler on ML Engineering and Safety
08 Apr 2026
Summary: 80,000 Hours Podcast — Paul Christiano on AI Alignment Solutions
08 Apr 2026
Summary: 80,000 Hours — Risks from Power-Seeking AI
08 Apr 2026
Summary: 80,000 Hours — Problem Profiles Overview
08 Apr 2026
Academic Papers Index — Bostrom, Parfit, MacAskill
08 Apr 2026
Summary: AI 2027 — A Scenario for Transformative AI
08 Apr 2026
AI in Context — YouTube Channel Videos
08 Apr 2026
Summary: Future Progress in Artificial Intelligence — A Survey of Expert Opinion
08 Apr 2026
Summary: Public Policy and Superintelligent AI -- A Vector Field Approach
08 Apr 2026
Summary: Existential Risk Prevention as Global Priority
08 Apr 2026
Summary: Existential Risks — Analyzing Human Extinction Scenarios and Related Hazards
08 Apr 2026
Summary: Optimal Timing for Superintelligence
08 Apr 2026
EA and AI Safety Books Reference
08 Apr 2026
EA and AI Safety Content Library — Master Inventory
08 Apr 2026
EA Forum — Key Posts on AI Safety and Longtermism
08 Apr 2026
Effective Altruism in the Age of AGI
08 Apr 2026
Key LessWrong and Alignment Forum Posts
08 Apr 2026
The Precipice Revisited — by Toby Ord
08 Apr 2026
Rationality: From AI to Zombies — by Eliezer Yudkowsky
08 Apr 2026
Summary: Situational Awareness Ch. I — From GPT-4 to AGI: Counting the OOMs
08 Apr 2026
Summary: Situational Awareness Ch. IIIa — Racing to the Trillion-Dollar Cluster
08 Apr 2026
Summary: Situational Awareness Ch. IIIb — Lock Down the Labs
08 Apr 2026
Summary: Situational Awareness — The Decade Ahead
07 Apr 2026
Summary — LLM Wiki by Andrej Karpathy
07 Apr 2026
Summary — NotebookLM vs. LLM Wiki: Two Approaches to AI Knowledge Management

Created with Quartz v0.1.0 © 2026

Suggest a source
Connect
Overview
About (proof of concept)
Email feedback
Made by IT for Humanity