Souradip Chakraborty — accepted AI-conference papers

2026

Direct Preference Optimization for Primitive-Enabled Hierarchical RL: A Bilevel Approach

Utsav Singh, Souradip Chakraborty, Wesley A. Suttle, Brian M. Sadler, Derrik E. Asher, Anit Kumar Sahu +3

ICLR 2026poster

Hierarchical reinforcement learning (HRL) enables agents to solve complex, long-horizon tasks by decomposing them into manageable sub-tasks. However, HRL methods face two fundamental challenges: (i) non-stationarity caused by the evolving lower-level policy during training, which destabilizes higher…

Cited by 0 Source ↗Scholar ↗

2026

Natural Language Actor–Critic Is Bilevel: Learning to Reason with Textual Feedback

Utsav Singh, Sidhaarth Murali, Souradip Chakraborty, Amrit Singh Bedi

ICML 2026poster

Reinforcement learning with verifiable rewards can improve LLM reasoning, but learning is sample-inefficient under sparse terminal rewards. Prior work mitigates this by adding natural language critiques, yet it typically treats critique generation as fixed or auxiliary, so correct-sounding feedback …

Cited by 0 Source ↗Scholar ↗

2026

Post-training Large Language Models for Diverse High-Quality Responses

Yilei Chen, Souradip Chakraborty, Lorenz Wolf, Ioannis Paschalidis, Aldo Pacchiano

ICLR 2026poster

Reinforcement learning has emerged as a popular method for post-training large language models (LLMs). While improving the model's performance on downstream tasks, it often reduces the model's output diversity, leading to narrow, canonical responses. Existing methods to enhance diversity are limited…

Cited by 0 Source ↗Scholar ↗

2026

Safety Recovery in Reasoning Models Is Only a Few Early Steering Steps Away

Soumya Suvra Ghosal, Souradip Chakraborty, Vaibhav Singh, Furong Huang, Dinesh Manocha, Amrit Singh Bedi

ICML 2026poster

Reinforcement learning (RL) based post-training for explicit chain-of-thought (e.g., GRPO) improves the reasoning ability of multimodal large-scale reasoning models (MLRMs). But recent evidence shows that it can simultaneously degrade safety alignment and increase jailbreak success rates. We propose…

Cited by 0 Source ↗Scholar ↗

2026

TEST-TIME SCALING IN DIFFUSION LLMS VIA HIDDEN SEMI-AUTOREGRESSIVE EXPERTS

Jihoon Lee, Hoyeon Moon, Kevin Zhai, Arun Kumar Chithanar, Anit Kumar Sahu, Soummya Kar +3

ICLR 2026poster

Diffusion-based large language models (dLLMs) are trained to model extreme flexibility/dependence in the data-distribution; however, how to best utilize this at inference time remains an open problem. In this work, we uncover an interesting property of these models: dLLMs {trained on textual data} i…

Cited by 0 Source ↗Scholar ↗

2025

A Technical Report on “Erasing the Invisible”: The 2024 NeurIPS Competition on Stress Testing Image Watermarks

Mucong Ding, Bang An, Tahseen Rabbani, Chenghao Deng, Anirudh Satheesh, Souradip Chakraborty +13

NeurIPS 2025poster

AI-generated images have become pervasive, raising critical concerns around content authenticity, intellectual property, and the spread of misinformation. Invisible watermarks offer a promising solution for identifying AI-generated images, preserving content provenance without degrading visual quali…

Cited by 0 Source ↗Scholar ↗

2025

Align-Pro: A Principled Approach to Prompt Optimization for LLM Alignment

Prashant Trivedi, Souradip Chakraborty, Avinash Reddy, Vaneet Aggarwal, Amrit Singh Bedi, George K. Atia

AAAI 2025technical

The alignment of large language models (LLMs) with human values is critical as these models become increasingly integrated into various societal and decision-making processes. Traditional methods, such as reinforcement learning from human feedback (RLHF), achieve alignment by fine-tuning model param…

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2025

Bounded Rationality for LLMs: Satisficing Alignment at Inference-Time

Mohamad Fares El Hajj Chehade, Soumya Suvra Ghosal, Souradip Chakraborty, Avinash Reddy, Dinesh Manocha, Hao Zhu +1

ICML 2025poster

Aligning large language models with humans is challenging due to the inherently multifaceted nature of preference feedback. While existing approaches typically frame this as a multi-objective optimization problem, they often overlook how humans actually make decisions. Research on bounded rationalit…

Cited by 0 Source ↗PDF ↗Scholar ↗

2025

Can Watermarking Large Language Models Prevent Copyrighted Text Generation and Hide Training Data?

Michael-Andrei Panaitescu-Liess, Zora Che, Bang An, Yuancheng Xu, Pankayaraj Pathmanathan, Souradip Chakraborty +3

AAAI 2025technical

Large Language Models (LLMs) have demonstrated impressive capabilities in generating diverse and contextually rich text. However, concerns regarding copyright infringement arise as LLMs may inadvertently produce copyrighted material. In this paper, we first investigate the effectiveness of watermark…

Cited by 4 Source ↗PDF ↗Scholar ↗

2025

Collab: Controlled Decoding using Mixture of Agents for LLM Alignment

Souradip Chakraborty, Sujay Bhatt, Udari Madhushani Sehwag, Soumya Suvra Ghosal, Jiahao Qiu, Mengdi Wang +4

ICLR 2025poster

Alignment of Large Language models (LLMs) is crucial for safe and trustworthy deployment in applications. Reinforcement learning from human feedback (RLHF) has emerged as an effective technique to align LLMs to human preferences, and broader utilities, but it requires updating billions of model para…

Cited by 1 Source ↗PDF ↗Scholar ↗

2025

Does Thinking More Always Help? Mirage of Test-Time Scaling in Reasoning Models

Soumya Suvra Ghosal, Souradip Chakraborty, Avinash Reddy, Yifu Lu, Mengdi Wang, Dinesh Manocha +3

NeurIPS 2025poster

Recent trends in test-time scaling for reasoning models (e.g., OpenAI o1, DeepSeek R1) have led to a popular belief that extending thinking traces using prompts like “Wait” or “Let me rethink” can improve performance. This raises a natural question: Does thinking more at test-time truly lead to bet…

Cited by 0 Source ↗Scholar ↗

2025

Immune: Improving Safety Against Jailbreaks in Multi-modal LLMs via Inference-Time Alignment

Soumya Suvra Ghosal, Souradip Chakraborty, Vaibhav Singh, Tianrui Guan, Mengdi Wang, Ahmad Beirami +4

CVPR 2025poster

With the widespread deployment of Multimodal Large Language Models (MLLMs) for visual-reasoning tasks, improving their safety has become crucial. Recent research indicates that despite training-time safety alignment, these models remain vulnerable to jailbreak attacks--carefully crafted image-prompt…

Cited by 3 Source ↗PDF ↗Scholar ↗

2025

Is Poisoning a Real Threat to DPO? Maybe More So Than You Think

Pankayaraj Pathmanathan, Souradip Chakraborty, Xiangyu Liu, Yongyuan Liang, Furong Huang

AAAI 2025technical

Recent advancements in Reinforcement Learning with Human Feedback (RLHF) have significantly impacted the alignment of Large Language Models (LLMs). The sensitivity of reinforcement learning algorithms such as Proximal Policy Optimization (PPO) has led to new line work on Direct Preference Optimizati…

Cited by 0 Source ↗PDF ↗Scholar ↗

2025

On the Global Optimality of Policy Gradient Methods in General Utility Reinforcement Learning

Anas Barakat, Souradip Chakraborty, Peihong Yu, Pratap Tokekar, Amrit Singh Bedi

NeurIPS 2025poster

Reinforcement learning with general utilities (RLGU) offers a unifying framework to capture several problems beyond standard expected returns, including imitation learning, pure exploration, and safe RL. Despite recent fundamental advances in the theoretical analysis of policy gradient (PG) methods…

Cited by 0 Source ↗Scholar ↗

2025

On the Vulnerability of LLM/VLM-Controlled Robotics

Xiyang Wu, Souradip Chakraborty, Ruiqi Xian, Jing Liang, Tianrui Guan, Fuxiao Liu +3

IROS 2025

In this work, we highlight vulnerabilities in robotic systems integrating large language models (LLMs) and vision-language models (VLMs) due to input modality sensitivities. While LLM/VLM-controlled robots show impressive performance across various tasks, their reliability under slight input variati

Cited by 13 Source ↗code ↗Scholar ↗

2025

Uncertainty-Aware Answer Selection for Improved Reasoning in Multi-LLM Systems

Aakriti Agrawal, Rohith Aralikatti, Anirudh Satheesh, Souradip Chakraborty, Amrit Singh Bedi, Furong Huang

EMNLP 2025

Large Language Models (LLMs) have demonstrated exceptional capabilities, yet selecting the most reliable response from multiple LLMs remains a challenge, particularly in resource-constrained settings. Existing approaches often depend on costly external verifiers, human evaluators, or self-consistenc

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2024

MaxMin-RLHF: Alignment with Diverse Human Preferences

Souradip Chakraborty, Jiahao Qiu, Hui Yuan, Alec Koppel, Dinesh Manocha, Furong Huang +2

ICML 2024poster

Reinforcement Learning from Human Feedback (RLHF) aligns language models to human preferences by employing a singular reward model derived from preference data. However, the single reward model overlooks the rich diversity of human preferences inherent in data collected from multiple users. In this…

Cited by 82 Source ↗PDF ↗Scholar ↗

2024

PARL: A Unified Framework for Policy Alignment in Reinforcement Learning from Human Feedback

Souradip Chakraborty, Amrit Bedi, Alec Koppel, Huazheng Wang, Dinesh Manocha, Mengdi Wang +1

ICLR 2024poster

We present a novel unified bilevel optimization-based framework, \textsf{PARL}, formulated to address the recently highlighted critical issue of policy alignment in reinforcement learning using utility or preference-based feedback. We identify a major gap within current algorithmic designs for solvi…

Cited by 29 Source ↗PDF ↗Scholar ↗

2024

Position: On the Possibilities of AI-Generated Text Detection

Souradip Chakraborty, Amrit Bedi, Sicheng Zhu, Bang An, Dinesh Manocha, Furong Huang

ICML 2024poster

Our study addresses the challenge of distinguishing human-written text from Large Language Model (LLM) outputs. We provide evidence that this differentiation is consistently feasible, except when human and machine text distributions are indistinguishable across their entire support. Employing inform…

Cited by 4 Source ↗PDF ↗Scholar ↗

2024

Rethinking Adversarial Policies: A Generalized Attack Formulation and Provable Defense in RL

Xiangyu Liu, Souradip Chakraborty, Yanchao Sun, Furong Huang

ICLR 2024poster

Most existing works focus on direct perturbations to the victim's state/action or the underlying transition dynamics to demonstrate the vulnerability of reinforcement learning agents to adversarial attacks. However, such direct manipulations may not be always realizable. In this paper, we consider…

Cited by 4 Source ↗PDF ↗Scholar ↗

2024

Transfer Q-star : Principled Decoding for LLM Alignment

Souradip Chakraborty, Soumya Suvra Ghosal, Ming Yin, Dinesh Manocha, Mengdi Wang, Amrit Bedi +1

NeurIPS 2024poster

Aligning foundation models is essential for their safe and trustworthy deployment. However, traditional fine-tuning methods are computationally intensive and require updating billions of model parameters. A promising alternative, alignment via decoding, adjusts the response distribution directly wit…

Cited by 20 Source ↗PDF ↗Scholar ↗

2023

Dealing with Sparse Rewards in Continuous Control Robotics via Heavy-Tailed Policy Optimization

Souradip Chakraborty, Amrit Singh Bedi, Kasun Weerakoon, Prithvi Poddar, Alec Koppel, Pratap Tokekar +1

ICRA 2023poster

In this paper, we present a novel Heavy-Tailed Stochastic Policy Gradient (HT-PSG) algorithm to deal with the challenges of sparse rewards in continuous control problems. Sparse rewards are common in continuous control robotics tasks such as manipulation and navigation and make the learning problem…

Cited by 3 Source ↗Scholar ↗

2023

Posterior Coreset Construction with Kernelized Stein Discrepancy for Model-Based Reinforcement Learning

Souradip Chakraborty, Amrit Singh Bedi, Pratap Tokekar, Alec Koppel, Brian Sadler, Furong Huang +1

AAAI 2023technical

Model-based approaches to reinforcement learning (MBRL) exhibit favorable performance in practice, but their theoretical guarantees in large spaces are mostly restricted to the setting when transition model is Gaussian or Lipschitz, and demands a posterior estimate whose representational complexity…

Cited by 11 Source ↗PDF ↗Scholar ↗

2023

STEERING : Stein Information Directed Exploration for Model-Based Reinforcement Learning

Souradip Chakraborty, Amrit Bedi, Alec Koppel, Mengdi Wang, Furong Huang, Dinesh Manocha

ICML 2023poster

Directed Exploration is a crucial challenge in reinforcement learning (RL), especially when rewards are sparse. Information-directed sampling (IDS), which optimizes the information ratio, seeks to do so by augmenting regret with information gain. However, estimating information gain is computational…

Cited by 8 Source ↗PDF ↗Scholar ↗

2022

HTRON: Efficient Outdoor Navigation with Sparse Rewards via Heavy Tailed Adaptive Reinforce Algorithm

Kasun Weerakoon, Souradip Chakraborty, Nare Karapetyan, Adarsh Jagan Sathyamoorthy, Amrit Bedi, Dinesh Manocha

CoRL 2022poster

We present a novel approach to improve the performance of deep reinforcement learning (DRL) based outdoor robot navigation systems. Most, existing DRL methods are based on carefully designed dense reward functions that learn the efficient behavior in an environment. We circumvent this issue by work…

Cited by 13 Source ↗Scholar ↗

2022

On the Hidden Biases of Policy Mirror Ascent in Continuous Action Spaces

Amrit Singh Bedi, Souradip Chakraborty, Anjaly Parayil, Brian M Sadler, Pratap Tokekar, Alec Koppel

ICML 2022spotlight

We focus on parameterized policy search for reinforcement learning over continuous action spaces. Typically, one assumes the score function associated with a policy is bounded, which {fails to hold even for Gaussian policies. } To properly address this issue, one must introduce an exploration tolera…

Cited by 20 Source ↗PDF ↗Scholar ↗

2020

BioMedBERT: A Pre-trained Biomedical Language Model for QA and IR

Souradip Chakraborty, Ekaba Bisong, Shweta Bhatt, Thomas Wagner, Riley Elliott, Francesco Mosconi

COLING 2020main

The SARS-CoV-2 (COVID-19) pandemic spotlighted the importance of moving quickly with biomedical research. However, as the number of biomedical research papers continue to increase, the task of finding relevant articles to answer pressing questions has become significant. In this work, we propose a t…

Cited by 61 Source ↗PDF ↗code ↗Scholar ↗