Author name cluster

Soham Dan

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

11 papers

2 author rows

TMLR Journal 2026 Journal Article

On Calibration of Multilingual Question Answering LLMs

Yahan Yang
Soham Dan
Dan Roth
Insup Lee

Multilingual pre-trained Large Language Models (LLMs) are incredibly effective at Question Answering (QA), a core task in Natural Language Understanding, achieving high accuracies on several multilingual benchmarks. However, little is known about how well their confidences are calibrated. In this paper, we comprehensively benchmark the calibration of several multilingual LLMs (MLLMs) on a variety of QA tasks. We perform extensive experiments, spanning encoder-only, encoder-decoder, and decoder-only QA models (size varying from 110M to 7B parameters) and diverse languages, including both high- and low-resource ones. We study different dimensions of calibration in in-distribution, out-of-distribution, and cross-lingual transfer settings, and investigate strategies to improve it, including post-hoc methods and regularized fine-tuning. For decoder-only LLMs such as LlaMa2, we additionally find that in-context learning improves confidence calibration on multilingual data. We also conduct several ablation experiments to study the effect of language distances, language corpus size, and model size on calibration, and how multilingual models compare with their monolingual counterparts for diverse tasks and languages. Our experiments suggest that the multilingual QA models are poorly calibrated for languages other than English and incorporating a small set of cheaply translated multilingual samples during fine-tuning/calibration effectively enhances the calibration performance.

PDF Details

TMLR Journal 2025 Journal Article

Exploring the Robustness of Language Models for Tabular Question Answering via Attention Analysis

Kushal Raj Bhandari
Sixue Xing
Soham Dan
Jianxi Gao

Large Language Models (LLMs), already shown to ace various unstructured text comprehension tasks, have also remarkably been shown to tackle table (structured) comprehension tasks without specific training. Building on earlier studies of LLMs for tabular tasks, we probe how in-context learning (ICL), model scale, instruction tuning, and domain bias affect Tabular QA (TQA) robustness by testing LLMs, under diverse augmentations and perturbations, on diverse domains: Wikipedia-based $\textbf{WTQ}$, financial $\textbf{TAT-QA}$, and scientific $\textbf{SCITAB}$. Although instruction tuning and larger, newer LLMs deliver stronger, more robust TQA performance, data contamination and reliability issues, especially on $\textbf{WTQ}$, remain unresolved. Through an in-depth attention analysis, we reveal a strong correlation between perturbation-induced shifts in attention dispersion and the drops in performance, with sensitivity peaking in the model's middle layers. We highlight the need for improved interpretable methodologies to develop more reliable LLMs for table comprehension. Through an in-depth attention analysis, we reveal a strong correlation between perturbation-induced shifts in attention dispersion and performance drops, with sensitivity peaking in the model's middle layers. Based on these findings, we argue for the development of structure-aware self-attention mechanisms and domain-adaptive processing techniques to improve the transparency, generalization, and real-world reliability of LLMs on tabular data.

PDF Details

TMLR Journal 2025 Journal Article

Large Language Model Confidence Estimation via Black-Box Access

Tejaswini Pedapati
Amit Dhurandhar
Soumya Ghosh
Soham Dan
Prasanna Sattigeri

Estimating uncertainty or confidence in the responses of a model can be significant in evaluating trust not only in the responses, but also in the model as a whole. In this paper, we explore the problem of estimating confidence for responses of large language models (LLMs) with simply black-box or query access to them. We propose a simple and extensible framework where, we engineer novel features and train a (interpretable) model (viz. logistic regression) on these features to estimate the confidence. We empirically demonstrate that our simple framework is effective in estimating confidence of Flan-ul2, Llama-13b, Mistral-7b and GPT-4 on four benchmark Q&A tasks as well as of Pegasus-large and BART-large on two benchmark summarization tasks with it surpassing baselines by even over 10% (on AU-ROC) in some cases. Additionally, our interpretable approach provides insight into features that are predictive of confidence, leading to the interesting and useful discovery that our confidence models built for one LLM generalize zero-shot across others on a given dataset.

PDF Details

ICLR Conference 2025 Conference Paper

Large Language Models can Become Strong Self-Detoxifiers

Ching-Yun Ko
Pin-Yu Chen
Payel Das
Youssef Mroueh
Soham Dan
Georgios Kollias
Subhajit Chaudhury
Tejaswini Pedapati

Reducing the likelihood of generating harmful and toxic output is an essential task when aligning large language models (LLMs). Existing methods mainly rely on training an external reward model (i.e., another language model) or fine-tuning the LLM using self-generated data to influence the outcome. In this paper, we show that LLMs have the capability of self-detoxification without external reward model learning or retraining of the LM. We propose \textit{Self-disciplined Autoregressive Sampling (SASA)}, a lightweight controlled decoding algorithm for toxicity reduction of LLMs. SASA leverages the contextual representations from an LLM to learn linear subspaces from labeled data characterizing toxic v.s. non-toxic output in analytical forms. When auto-completing a response token-by-token, SASA dynamically tracks the margin of the current output to steer the generation away from the toxic subspace, by adjusting the autoregressive sampling strategy. Evaluated on LLMs of different scale and nature, namely Llama-3.1-Instruct (8B), Llama-2 (7B), and GPT2-L models with the RealToxicityPrompts, BOLD, and AttaQ benchmarks, SASA markedly enhances the quality of the generated sentences relative to the original models and attains comparable performance to state-of-the-art detoxification techniques, significantly reducing the toxicity level by only using the LLM's internal representations.

Details

TMLR Journal 2025 Journal Article

On the Utility of Existing Fine-Tuned Models on Data-Scarce Domains

Md Ibrahim Ibne Alam
Parikshit Ram
Soham Dan
Horst Samulowitz
Koushik Kar

Large Language Models (LLMs) have been observed to perform well on a wide range of downstream tasks when fine-tuned on domain-specific data. However, such data may not be readily available in many applications, motivating zero-shot or few-shot approaches using existing domain or task adjacent (fine-tuned) models, which we call DAFT. While several fine-tuned models for various tasks are available, finding one appropriate DAFT model for a given task is often not straight forward. In this paper, we explore different utilization techniques of these existing DAFT models for data-scarce problems, i.e., tasks for which data is not available or limited. We observe that for zero-shot problems, ensembling of DAFT models provides an accuracy performance close to that of the single best model. With few-shot problems (few data from target domain available), this performance can be improved further by picking or putting more weights to the DAFT models that are expected to perform better on the target task.

PDF Details

AAAI Conference 2024 Conference Paper

Generalized Planning in PDDL Domains with Pretrained Large Language Models

Tom Silver
Soham Dan
Kavitha Srinivas
Joshua B. Tenenbaum
Leslie Kaelbling
Michael Katz

Recent work has considered whether large language models (LLMs) can function as planners: given a task, generate a plan. We investigate whether LLMs can serve as generalized planners: given a domain and training tasks, generate a program that efficiently produces plans for other tasks in the domain. In particular, we consider PDDL domains and use GPT-4 to synthesize Python programs. We also consider (1) Chain-of-Thought (CoT) summarization, where the LLM is prompted to summarize the domain and propose a strategy in words before synthesizing the program; and (2) automated debugging, where the program is validated with respect to the training tasks, and in case of errors, the LLM is re-prompted with four types of feedback. We evaluate this approach in seven PDDL domains and compare it to four ablations and four baselines. Overall, we find that GPT-4 is a surprisingly powerful generalized planner. We also conclude that automated debugging is very important, that CoT summarization has non-uniform impact, that GPT-4 is far superior to GPT-3.5, and that just two training tasks are often sufficient for strong generalization.

PDF Details DOI

ICML Conference 2024 Conference Paper

Larimar: Large Language Models with Episodic Memory Control

Payel Das
Subhajit Chaudhury
Elliot Nelson
Igor Melnyk
Sarathkrishna Swaminathan
Sihui Dai
Aurélie C. Lozano
Georgios Kollias

Efficient and accurate updating of knowledge stored in Large Language Models (LLMs) is one of the most pressing research challenges today. This paper presents Larimar - a novel, brain-inspired architecture for enhancing LLMs with a distributed episodic memory. Larimar’s memory allows for dynamic, one-shot updates of knowledge without the need for computationally expensive re-training or fine-tuning. Experimental results on multiple fact editing benchmarks demonstrate that Larimar attains accuracy comparable to most competitive baselines, even in the challenging sequential editing setup, but also excels in speed—yielding speed-ups of 8-10x depending on the base LLM —as well as flexibility due to the proposed architecture being simple, LLM-agnostic, and hence general. We further provide mechanisms for selective fact forgetting, information leakage prevention, and input context length generalization with Larimar and show their effectiveness. Our code is available at https: //github. com/IBM/larimar.

Details

ICLR Conference 2024 Conference Paper

On the generalization capacity of neural networks during generic multimodal reasoning

Takuya Ito
Soham Dan
Mattia Rigotti
James R. Kozloski
Murray Campbell

The advent of the Transformer has led to the development of large language models (LLM), which appear to demonstrate human-like capabilities. To assess the generality of this class of models and a variety of other base neural network architectures to multimodal domains, we evaluated and compared their capacity for multimodal generalization. We introduce a multimodal question-answer benchmark to evaluate three specific types of out-of-distribution (OOD) generalization performance: distractor generalization (generalization in the presence of distractors), systematic compositional generalization (generalization to new task permutations), and productive compositional generalization (generalization to more complex tasks with deeper dependencies). While we found that most architectures faired poorly on most forms of generalization (e.g., RNNs and standard Transformers), models that leveraged cross-attention mechanisms between input domains, such as the Perceiver, fared better. Our positive results demonstrate that for multimodal distractor and systematic generalization, cross-attention is an important mechanism to integrate multiple sources of information. On the other hand, all architectures failed in productive generalization, suggesting fundamental limitations of existing architectures for specific types of multimodal OOD generalization. These results demonstrate the strengths and limitations of specific architectural components underlying modern neural models for multimodal reasoning. Finally, we provide *Generic COG* (gCOG), a configurable benchmark with several multimodal generalization splits, for future studies to explore.

Details

PRL Workshop 2023 Workshop Paper

Generalized Planning in PDDL Domains with Pretrained Large Language Models

Tom Silver
Soham Dan
Kavitha Srinivas
Joshua B. Tenenbaum
Leslie Pack Kaelbling
Michael Katz

PDF Details

ICML Conference 2022 Conference Paper

Understanding Robust Generalization in Learning Regular Languages

Soham Dan
Osbert Bastani
Dan Roth 0001

A key feature of human intelligence is the ability to generalize beyond the training distribution, for instance, parsing longer sentences than seen in the past. Currently, deep neural networks struggle to generalize robustly to such shifts in the data distribution. We study robust generalization in the context of using recurrent neural networks (RNNs) to learn regular languages. We hypothesize that standard end-to-end modeling strategies cannot generalize well to systematic distribution shifts and propose a compositional strategy to address this. We compare an end-to-end strategy that maps strings to labels with a compositional strategy that predicts the structure of the deterministic finite state automaton (DFA) that accepts the regular language. We theoretically prove that the compositional strategy generalizes significantly better than the end-to-end strategy. In our experiments, we implement the compositional strategy via an auxiliary task where the goal is to predict the intermediate states visited by the DFA when parsing a string. Our empirical results support our hypothesis, showing that auxiliary tasks can enable robust generalization. Interestingly, the end-to-end RNN generalizes significantly better than the theoretical lower bound, suggesting that it is able to achieve atleast some degree of robust generalization.

Details

ICML Conference 2020 Conference Paper

Goodness-of-Fit Tests for Inhomogeneous Random Graphs

Soham Dan
Bhaswar B. Bhattacharya

Hypothesis testing of random networks is an emerging area of modern research, especially in the high-dimensional regime, where the number of samples is smaller or comparable to the size of the graph. In this paper we consider the goodness-of-fit testing problem for large inhomogeneous random (IER) graphs, where given a (known) reference symmetric matrix $Q \in [0, 1]^{n \times n}$ and $m$ independent samples from an IER graph given by an unknown symmetric matrix $P \in [0, 1]^{n \times n}$, the goal is to test the hypothesis $P=Q$ versus $||P-Q|| \geq \varepsilon$, where $||\cdot||$ is some specified norm on symmetric matrices. Building on recent related work on two-sample testing for IER graphs, we derive the optimal minimax sample complexities for the goodness-of-fit problem in various natural norms, such as the Frobenius norm and the operator norm. We also propose practical implementations of natural test statistics, using their asymptotic distributions and through the parametric bootstrap. We compare the performances of the different tests in simulations, and show that the proposed tests outperform the baseline tests across various natural random graphs models.

Details