2 citations · 4 across the 13 of their papers we have counts for
9 papers · 1 filter
GitSwarm: Decentralized Compounding Inference
Vedant Shah, Ankur Samanta, Paras Dahal +11
Long-horizon problem solving and scientific research require computation to accumulate across successive attempts. Partial solutions, experimental findings, and unsuccessful approa…
Learning What to Investigate Next: Meta-Reasoning for Long-Horizon Research Agents
Ankur Samanta, Yonathan Efroni, Paul Sajda +2
Long-horizon research agents must decide both how to investigate and what to investigate next as evidence accumulates. This is hard to learn because such decisions are sparse in lo…
Off-Context GRPO: Learning to Reason on Hard Problems using Privileged Information
Priyank Agrawal, Ankur Samanta, Shervin Ghasemlou +5
Reinforcement learning with verifiable rewards (RLVR) improves reasoning in large language models. Yet, typical RLVR approaches fail on difficult problems: when a model cannot gene…
Frustratingly Simple Black-Box Adaptation of Language Models via Logit Bias
Ofek I. Cohen, Lior Shani, Aviv Rosenberg +3
Many organizations aim to adapt language models for internal use, both to improve performance on domain-specific tasks and to address privacy concerns around sensitive data. Howeve…
BayesBench: Evaluating LLM Belief Trajectories Under Multi-Turn Evidence Accumulation
Ankur Samanta, Akshayaa Magesh, Tal Lancewicki +7
Large language models (LLMs) are typically deployed in multi-turn conversations, where each turn provides new evidence that should reduce epistemic uncertainty about their environm…
Credit Assignment with Resets in Language Model Reasoning
Ankur Samanta, Akshayaa Magesh, Ayush Jain +7
Contemporary reinforcement learning with verifiable reward methods post-train language models on multi-step reasoning by assigning a single outcome reward uniformly across all toke…
Hack-Verifiable Environments: Towards Evaluating Reward Hacking at Scale
Amit Roth, Ankur Samanta, Matan Halevy +2
Aligning autonomous agents with human intent remains a central challenge in modern AI. A key manifestation of this challenge is reward hacking, whereby agents appear successful und…
Gaze patterns predict preference and confidence in pairwise AI image evaluation
Nikolas Papadopoulos, Shreenithi Navaneethan, Sheng Bai +2
Preference learning methods, such as Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimization (DPO), rely on pairwise human judgments, yet little is kno…
Structure Enables Effective Self-Localization of Errors in LLMs
Ankur Samanta, Akshayaa Magesh, Ayush Jain +8
Self-correction in language models remains elusive. In this work, we explore whether language models can explicitly localize errors in incorrect reasoning, as a path toward buildin…