activity
20192026
most citedFederated Learning with Personalization Layers

523 citations · 531 across the 8 of their papers we have counts for

collaborators

9 papers

cs.LG2026

Don't Mask the Environment: Observation Supervision Changes How Agents Explore Under RL

Juzheng Zhang, Disha Makhija, Manoj Ghuhan Arivazhagan +2

Agent trajectories record what an agent does and what happens next. Yet standard supervised fine-tuning (SFT) applies loss only to agent-authored action tokens, using environment o…

cs.CL2026

Probing the Prompt KV Cache: Where It Becomes Dispensable

Vinayshekhar Bannihatti Kumar, Manoj Ghuhan Arivazhagan, Disha Makhija +1

Prior KV cache compression schemes empirically demonstrate that the prompt cache is partially redundant during decoding, dropping or summarising entries with little accuracy loss.…

cs.CL2026

Syntax Without Semantics: Teaching Large Language Models to Code in an Unseen Language

Vinayshekhar Bannihatti Kumar, Disha Makhija, Manoj Ghuhan Arivazhagan +1

Large language models (LLMs) achieve high pass rates on code generation benchmarks, yet whether they can transfer this ability to languages absent from pretraining remains poorly u…

cs.CL2025

When Facts Change: Probing LLMs on Evolving Knowledge with evolveQA

Nishanth Sridhar Nakshatri, Shamik Roy, Manoj Ghuhan Arivazhagan +3

LLMs often fail to handle temporal knowledge conflicts--contradictions arising when facts evolve over time within their training data. Existing studies evaluate this phenomenon thr…

cs.LG2025★ 1 cited

Neural Breadcrumbs: Membership Inference Attacks on LLMs Through Hidden State and Attention Pattern Analysis

Disha Makhija, Manoj Ghuhan Arivazhagan, Vinayshekhar Bannihatti Kumar +1

Membership inference attacks (MIAs) reveal whether specific data was used to train machine learning models, serving as important tools for privacy auditing and compliance assessmen…

cs.CV2022

SALAD: Source-free Active Label-Agnostic Domain Adaptation for Classification, Segmentation and Detection

Divya Kothandaraman, Sumit Shekhar, Abhilasha Sancheti +3

We present a novel method, SALAD, for the challenging vision task of adapting a pre-trained "source" domain network to a "target" domain, with a small budget for annotation in the…