8 papers
A False Average: Pooled CoT-Monitor Accuracy Conceals a Reasoning-Dependent Fragility
Shikhar Shiromani, Leo Richter, Noah Siegel
Chain-of-thought (CoT) monitors are reported by their aggregate accuracy on a pool of reward hacks. We show that this number is a false average. On Terminal Wrench, about 77% of ha…
Visuals Lie, Consistency Speaks: Disentangling Spatial Attention from Reliability in Vision-Language Models
Logan Mann, Yi Xia, Ajit Saravanan +6
Multimodal Foundation Models are increasingly used as reasoning agents, making reliability, knowing when a model may hallucinate, critical. A common intuition, which we call the At…
Where Reliability Lives in Vision-Language Models: A Mechanistic Study of Attention, Hidden States, and Causal Circuits
Logan Mann, Ajit Saravanan, Ishan Dave +4
A pervasive intuition holds that vision-language models (VLMs) are most trustworthy when their attention maps look sharp: concentrated attention on the queried region should imply…
Linear Predictability of Attention Heads in Large Language Models
Khalid Shaikh, Asmit Kumar Singh, Rebecca Christopher Dsouza +1
Large language model (LLM) inference is increasingly bottlenecked by the Key-Value (KV) cache, yet the fine-grained structure of attention-head activations remains poorly understoo…
ProMoral-Bench: Evaluating Prompting Strategies for Moral Reasoning and Safety in LLMs
Rohan Subramanian Thomas, Shikhar Shiromani, Abdullah Chaudhry +4
Prompt design significantly impacts the moral competence and safety alignment of large language models (LLMs), yet empirical comparisons remain fragmented across datasets and model…
SAGE: Agentic Framework for Interpretable and Clinically Translatable Computational Pathology Biomarker Discovery
Sahar Almahfouz Nasser, Juan Francisco Pesantez Borja, Jincheng Liu +15
Engineered image-based biomarkers offer a clinically interpretable alternative to black-box AI in computational pathology, yet their discovery remains largely intuition-driven, gui…
The Hypocrisy Gap: Quantifying Divergence Between Internal Belief and Chain-of-Thought Explanation via Sparse Autoencoders
Shikhar Shiromani, Archie Chaudhury, Sri Pranav Kunda
Large Language Models (LLMs) frequently exhibit unfaithful behavior, producing a final answer that differs significantly from their internal chain of thought (CoT) reasoning in ord…
COMPASS: Context-Modulated PID Attention Steering System for Hallucination Mitigation
Kenji Sahay, Snigdha Pandya, Rohan Nagale +4
Large language models (LLMs) often generate fluent but factually incorrect statements despite having access to relevant evidence, a failure mode rooted in how they allocate attenti…