Vinija Jain

Swiss-Knife: A Framework for Reconfigurable Externalised Multi-Objective Alignment at Decode Time

Agnibh Karmakar, Mayur Parvatikar, Shreyash Dhoot +5

Decode-time alignment methods steer a frozen language model by scoring candidate continuations with an external reward and selecting the maximiser. We argue that this shared design…

RECOM: A Validity Discrimination Tradeoff in Automatic Metrics for Open Ended Reddit Question Answering

Pushwitha Krishnappa, Amit Das, Vinija Jain +2

Automatic metrics are the default for evaluating LLM-generated text, yet a metric is quietly asked to do two jobs: tell genuine content alignment from surface coincidence (validity…

MENTIS: What Belief Changes Under Alignment? Measuring Multi-Scale Latent Torsion in Language Models

Partha Pratim Saha, Samarth Raina, Mayur Parvatikar +4

Preference alignment has substantially improved the observable behavior of large language models, yet it remains unclear what alignment changes internally. Aligned systems still fa…

Linear Probes Detect Task Format, Not Reasoning Mode in Language Model Hidden States

Subramanyam Sahoo, Vinija Jain, Aman Chadha +1

Linear probing of large language model (LLM) hidden states is widely used to claim that models learn distinct representations for different reasoning types. We test this by probing…

Findings of the Counter Turing Test: AI-Generated Text Detection

Rajarshi Roy, Gurpreet Singh, Ashhar Aziz +16

The growing capability of large language models to produce fluent, contextually coherent text has created mounting pressure on the systems and institutions responsible for ensuring…

Personality Shapes Gender Bias in Persona-Conditioned LLM Narratives Across English and Hindi: An Empirical Investigation

Tanay Kumar, Shreya Gautam, Aman Chadha +2

Large Language Models (LLMs) are increasingly deployed in persona-driven applications such as education, customer service, and social platforms, where models are prompted to adopt…

Neural FOXP2 -- Language Specific Neuron Steering for Targeted Language Improvement in LLMs

Anusa Saha, Tanmay Joshi, Vinija Jain +2

LLMs are multilingual by training, yet their lingua franca is often English, reflecting English language dominance in pretraining. Other languages remain in parametric memory but a…

Assessing LLM Reliability on Temporally Recent Open-Domain Questions

Pushwitha Krishnappa, Amit Das, Vinija Jain +2

Large Language Models (LLMs) are increasingly deployed for open-domain question answering, yet their alignment with human perspectives on temporally recent information remains unde…

Catch Me If You Can: How Smaller Reasoning Models Pretend to Reason with Mathematical Fidelity

Subramanyam Sahoo, Vinija Jain, Saanidhya Vats +4

Current evaluation of mathematical reasoning in language models relies primarily on answer accuracy, potentially masking fundamental failures in logical computation. We introduce a…

A Comprehensive Dataset for Human vs. AI Generated Text Detection

Rajarshi Roy, Gurpreet Singh, Ashhar Aziz +17

The rapid advancement of large language models (LLMs) has led to increasingly human-like AI-generated text, raising concerns about content authenticity, misinformation, and trustwo…

AMBEDKAR-A Multi-level Bias Elimination through a Decoding Approach with Knowledge Augmentation for Robust Constitutional Alignment of Language Models

Snehasis Mukhopadhyay, Aryan Kasat, Shivam Dubey +5

Large Language Models (LLMs) can inadvertently reflect societal biases present in their training data, leading to harmful or prejudiced outputs. In the Indian context, our empirica…

Investigating Hallucination in Conversations for Low Resource Languages

Amit Das, Md. Najib Hasan, Souvika Sarkar +7

Large Language Models (LLMs) have demonstrated remarkable proficiency in generating text that closely resemble human writing. However, they often generate factually incorrect state…

RADIANT: Retrieval AugmenteD entIty-context AligNmenT -- Introducing RAG-ability and Entity-Context Divergence

Vipula Rawte, Rajarshi Roy, Gurpreet Singh +11

As Large Language Models (LLMs) continue to advance, Retrieval-Augmented Generation (RAG) has emerged as a vital technique to enhance factual accuracy by integrating external knowl…

QuickSilver -- Speeding up LLM Inference through Dynamic Token Halting, KV Skipping, Contextual Token Fusion, and Adaptive Matryoshka Quantization

Danush Khanna, Aditya Kumar Guru, Srivarshinee Sridhar +7

Inference accounts for the majority of latency and energy consumption in large language model (LLM) deployments, often exceeding 90% of total cost. While training-time efficiency h…

Alignment Quality Index (AQI) : Beyond Refusals: AQI as an Intrinsic Alignment Diagnostic via Latent Geometry, Cluster Divergence, and Layer wise Pooled Representations

Abhilekh Borah, Chhavi Sharma, Danush Khanna +12

Alignment is no longer a luxury, it is a necessity. As large language models (LLMs) enter high-stakes domains like education, healthcare, governance, and law, their behavior must r…

AdversariaL attacK sAfety aLIgnment(ALKALI): Safeguarding LLMs through GRACE: Geometric Representation-Aware Contrastive Enhancement- Introducing Adversarial Vulnerability Quality Index (AVQI)

Danush Khanna, Gurucharan Marthi Krishna Kumar, Basab Ghosh +5

Adversarial threats against LLMs are escalating faster than current defenses can adapt. We expose a critical geometric blind spot in alignment: adversarial prompts exploit latent c…

cs.CL2025★ 1 cited

Multilingual State Space Models for Structured Question Answering in Indic Languages

Arpita Vats, Rahul Raja, Mrinal Mathur +2

The diversity and complexity of Indic languages present unique challenges for natural language processing (NLP) tasks, particularly in the domain of question answering (QA).To addr…

IndicMMLU-Pro: Benchmarking Indic Large Language Models on Multi-Task Language Understanding

Sankalp KJ, Ashutosh Kumar, Laxmaan Balaji +4

Known by more than 1.5 billion people in the Indian subcontinent, Indic languages present unique challenges and opportunities for natural language processing (NLP) research due to…

LLMsAgainstHate @ NLU of Devanagari Script Languages 2025: Hate Speech Detection and Target Identification in Devanagari Languages via Parameter Efficient Fine-Tuning of LLMs

Rushendra Sidibomma, Pransh Patwa, Parth Patwa +3

The detection of hate speech has become increasingly important in combating online hostility and its real-world consequences. Despite recent advancements, there is limited research…

SKETCH: Structured Knowledge Enhanced Text Comprehension for Holistic Retrieval

Aakash Mahalingam, Vinesh Kumar Gande, Aman Chadha +2

Retrieval-Augmented Generation (RAG) systems have become pivotal in leveraging vast corpora to generate informed and contextually relevant responses, notably reducing hallucination…

KnowledgePrompts: Exploring the Abilities of Large Language Models to Solve Proportional Analogies via Knowledge-Enhanced Prompting

Thilini Wijesiriwardene, Ruwan Wickramarachchi, Sreeram Vennam +5

Making analogies is fundamental to cognition. Proportional analogies, which consist of four terms, are often used to assess linguistic and cognitive abilities. For instance, comple…