Who Evaluates AI's Social Impacts? Mapping Coverage and Gaps in First and Third Party Evaluations
Paper
• 2511.05613 • Published
Subramanyam Sahoo
SahoobhAI
AI & ML interests
RL, AI Safety and AI Alignment
Recent Activity
authored
a paper
about 2 hours ago
Who Evaluates AI's Social Impacts? Mapping Coverage and Gaps in First and Third Party Evaluations authored
a paper
about 2 hours ago
Position: The Complexity of Perfect AI Alignment -- Formalizing the RLHF Trilemma authored
a paper
about 2 hours ago
Catch Me If You Can: How Smaller Reasoning Models Pretend to Reason with Mathematical Fidelity Organizations
authored 8 papers about 2 hours ago
Position: The Complexity of Perfect AI Alignment -- Formalizing the RLHF Trilemma
Paper
• 2511.19504 • Published
• 2
Catch Me If You Can: How Smaller Reasoning Models Pretend to Reason with Mathematical Fidelity
Paper
• 2512.00552 • Published
When Shallow Wins: Silent Failures and the Depth-Accuracy Paradox in Latent Reasoning
Paper
• 2603.03475 • Published
I Can't Believe It's Not Robust: Catastrophic Collapse of Safety Classifiers under Embedding Drift
Paper
• 2603.01297 • Published
Dial E for Ethical Enforcement: institutional VETO power as a governance primitive
Paper
• 2603.00617 • Published
SAHOO: Safeguarded Alignment for High-Order Optimization Objectives in Recursive Self-Improvement
Paper
• 2603.06333 • Published
• 1
The Reasoning Trap -- Logical Reasoning as a Mechanistic Pathway to Situational Awareness
Paper
• 2603.09200 • Published
• 5
upvoted a paper about 19 hours ago
liked a model 5 months ago
upvoted an article over 1 year ago
Article
Let's talk about LLM evaluation
•
207