Papers by Subhajit Chaudhury

19 papers
X-FACTOR: A Cross-metric Evaluation of Factual Correctness in Abstractive Summarization (2022.emnlp-main)

Copied to clipboard

Challenge: Abstractive summarization models produce factually inconsistent summaries that are not supported by the original article.
Approach: They propose a fact-aware filtering mechanism that improves the factuality of abstractive summarization models.
Outcome: The proposed method improves the quality of training data and the factuality of generated summaries.
Bootstrapped Q-learning with Context Relevant Observation Pruning to Generalize in Text-based Games (2020.emnlp-main)

Copied to clipboard

Challenge: Reinforcement Learning methods for text-based games fail to generalize on unseen games, especially in small data regimes.
Approach: They propose a Context Relevant Episodic State Truncation method for irrelevant token removal in observation text for improved generalization.
Outcome: The proposed method shows that it can generalize on unseen games using 10x-20x fewer training games compared to previous state-of-the-art methods despite requiring fewer number of training episodes.
EXPLORER: Exploration-guided Reasoning for Textual Reinforcement Learning (2024.eacl-long)

Copied to clipboard

Challenge: Text-based games (TBGs) combine natural language understanding with reasoning.
Approach: They propose an exploration-guided reasoning agent for textual reinforcement learning that integrates natural language with reasoning.
Outcome: The proposed agent outperforms baseline agents on TWG and TWC games.
ImReasoner: Improving Memory-based Language Models for Reasoning-in-a-Haystack Tasks (2026.acl-long)

Copied to clipboard

Challenge: despite advances, large language models exhibit brittleness on tasks that require multi-step reasoning over long contexts.
Approach: They propose to explicitly encode contexts as ordered memory and perform iterative retrieval to construct reasoning chains.
Outcome: The proposed frameworks fail to show emergent reasoning generalization in a weakly supervised scenario . the proposed framework is based on a synthetic benchmark to stress-test the models .
ZoomR: Memory Efficient Reasoning through Multi-Granularity Key Value Retrieval (2026.acl-long)

Copied to clipboard

Challenge: Large language models generate long chain of thoughts but memory footprint grows with output length . prior work on KV cache optimization focused on compressing long input context .
Approach: They propose a new approach that compresses verbose reasoning thoughts into summaries . they use a dynamic KV cache selection policy that leverages these summary keys .
Outcome: The proposed approach reduces memory usage while avoiding full-cache attention at each step.
Answering the Wrong Question: Reasoning Trace Inversion for Abstention in LLMs (2026.acl-long)

Copied to clipboard

Challenge: Large Language Models have been shown to have worse abstention abilities than reasoning models . a new class of abstraction methods is developed to improve absttention performance .
Approach: They propose a class of abstention methods that generate reasoning trace and reconstruct most likely query from it.
Outcome: The proposed method beats baselines in 33 out of 36 settings.
LOA: Logical Optimal Actions for Text-based Interaction Games (2021.acl-demo)

Copied to clipboard

Challenge: et al., 2019) have proposed a neuro-symbolic approach for reinforcement learning in non-simultaneous environments.
Approach: They propose an action decision architecture with a neuro-symbolic framework for natural language interaction games.
Outcome: The proposed framework provides an open-source implementation in Python for the reinforcement learning environment to facilitate an experiment for studying neuro-symbolic agents.
MISMATCH: Fine-grained Evaluation of Machine-generated Text with Mismatch Error Types (2023.findings-acl)

Copied to clipboard

Challenge: Existing evaluation metrics for machine text are inadequate to capture quality of text . a recent study has focused on task-specific evaluation metrics or on properties of machine-generated text based on mismatch errors .
Approach: They propose a new evaluation scheme based on fine-grained mismatch errors . they propose 13 mismatch error types to guide the model for better prediction of human judgments .
Outcome: The proposed evaluation scheme is based on mismatch errors in 7 NLP tasks . the mismatch error types guide the model for better prediction of human judgments .
Language-based General Action Template for Reinforcement Learning Agents (2021.findings-acl)

Copied to clipboard

Challenge: Prior knowledge is important in decision-making, and humans preserve it in the form of natural language (NL).
Approach: They propose an environmentagnostic action framework that incorporates prior knowledge into decision-making . they propose to use general semantic schemes to facilitate agent in finding plausible actions .
Outcome: The proposed agent performs better than agents that rely on gamespecific actions.
On the Effects of Fine-tuning Language Models for Text-Based Reinforcement Learning (2025.coling-main)

Copied to clipboard

Challenge: Text-based reinforcement learning is a form of interactive fiction where players manipulate the environment using text and admissible actions in natural language.
Approach: They show that rich semantic understanding leads to efficient training of text-based RL agents . they also show that semantic degeneration occurs when LMs are inappropriately fine-tuned .
Outcome: The results suggest that semantic understanding is not important for the task . they also show that fine-tuning language models can degenerate the agent's performance .
Neuro-Symbolic Reinforcement Learning with First-Order Logic (2021.emnlp-main)

Copied to clipboard

Challenge: Existing deep reinforcement learning methods require many trials before convergence and no direct interpretability of trained policies is provided.
Approach: They propose a novel RL method which can learn symbolic and interpretable rules in their differentiable network.
Outcome: The proposed method can learn symbolic and interpretable rules in their differentiable network.
Self-Supervised Rule Learning to Link Text Segments to Relational Elements of Structured Knowledge (2023.findings-emnlp)

Copied to clipboard

Challenge: Various approaches have been tried to map predicate components of a natural language (NL) text segment onto their corresponding predicates within a knowledge base (KB).
Approach: They propose a neuro-symbolic approach to self-learn rules that serve as interpretable knowledge to perform relation linking in knowledge base question answering systems.
Outcome: The proposed approach achieves an average performance gain of 17% on CLUTRR and relation linking in a KBQA system.
Granite Guardian: Comprehensive LLM Safeguarding (2025.naacl-industry)

Copied to clipboard

Challenge: a suite of advanced models is designed to detect and mitigate risks associated with prompts and responses.
Approach: a team of researchers develop a model family to detect and mitigate risks associated with prompts and responses. the model family is based on the Granite 3.0 language models.
Outcome: a new model family is designed to detect and mitigate risks associated with prompts and responses.
Learning Symbolic Rules over Abstract Meaning Representations for Textual Reinforcement Learning (2023.acl-long)

Copied to clipboard

Challenge: Existing text-based reinforcement learning agents use embeddings as representations for observation and are fed to an action scorer for predicting the next action.
Approach: They propose a novel neurosymbolic agent that combines a semantic parser and a rule induction system to learn interpretable rules as policies.
Outcome: The proposed method outperforms deep learning-based methods on established text-based game benchmarks on unobserved games and on unseen games.
EpMAN: Episodic Memory AttentioN for Generalizing to Longer Contexts (2025.acl-long)

Copied to clipboard

Challenge: Recent advances in Large Language Models (LLMs) have yielded impressive successes on many language tasks, but efficient processing of long contexts remains a significant challenge.
Approach: They propose a method for processing long contexts in an episodic memory module while holistically attending to semantically-relevant context chunks.
Outcome: The proposed method outperforms baseline decoders on multiple long-context recall and question-answering benchmarks on 16k to 256k tokens.
Neuro-Symbolic Approaches for Text-Based Policy Learning (2021.emnlp-main)

Copied to clipboard

Challenge: Text-based games are important testbeds for reinforcement learning in the natural language domain.
Approach: They propose a method that learns interpretable action policy rules from symbolic abstractions of textual observations for improved generalization.
Outcome: The proposed method outperforms existing methods in RL using 5-10x fewer training games.
API-BLEND: A Comprehensive Corpora for Training and Benchmarking API LLMs (2024.acl-long)

Copied to clipboard

Challenge: Existing methods to train and test large language models that involve calls to tools and APIs are lacking.
Approach: They propose a large corpora for training and systematic testing of tool-augmented LLMs.
Outcome: The proposed datasets mimic real-world scenarios involving API-tasks and slot filling.
A Grounded Preference Model for LLM Alignment (2024.findings-acl)

Copied to clipboard

Challenge: Large Language Models (LLMs) suffer from factual inconsistency and hallucination despite recent advances . training a preference model requires substantial human annotation, which is expensive and labor-intensive.
Approach: They propose to generate synthetic grounded preference data and train a Grounded Preference Model to assess the overall quality of grounded responses.
Outcome: The proposed model can generate much better grounded responses as judged by GPT4 and achieves the TRUE faithfulness Benchmark.
Laziness Is a Virtue When It Comes to Compositionality in Neural Semantic Parsing (2023.acl-long)

Copied to clipboard

Challenge: Compositional generalization is a key feature of human intelligence and has been identified as a major point of weakness in neural methods for semantic parsing.
Approach: They propose a neural parsing generation method that constructs logical forms from the bottom up, beginning from the logical form’s leaves.
Outcome: The proposed method outperforms general-purpose parsers on a CFQ dataset and two other Text-to-SQL datasets while also being competitive with parser that have been tailored to each task.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations