AI.info

Subramanyam Sahoo

Explore Subramanyam Sahoo on AI.info.

• The Reasoning Trap -- Logical Reasoning as a Mechanistic Pathway to Situational Awareness
• SAHOO: Safeguarded Alignment for High-Order Optimization Objectives in Recursive Self-Improvement
• The Controllability Trap: A Governance Framework for Military AI Agents
• Policy myopia as a mechanism of gradual disempowerment in Post-AGI governance, Circa 2049
• Dial E for Ethical Enforcement: institutional VETO power as a governance primitive
• Position: The Complexity of Perfect AI Alignment -- Formalizing the RLHF Trilemma
• The Horcrux: Mechanistically Interpretable Task Decomposition for Detecting and Mitigating Reward Hacking in Embodied AI Systems
• The Good, The Bad, and The Hybrid: A Reward Structure Showdown in Reasoning Models Training
• The Last Vote: A Multi-Stakeholder Framework for Language Model Governance
• Who Evaluates AI's Social Impacts? Mapping Coverage and Gaps in First and Third Party Evaluations