SeT-LLM @ KDD 2026
Accepted (Poster)
A binary QA model can answer a question correctly and still give the same label to its negation. We test whether Mutual Information Transfer Regularization, which discourages adjacent transformer layers from making redundant residual contributions, reduces those contradictions. Across an estimator screen, three backbones, a weight sweep and a five-seed replication, the answer is a clean negative: the gains are small, unstable, and not competitive with direct supervision on negated examples.
Forecast @ ICML 2026
Accepted
James Begin, Brendan Gho, Suman Muppavarapu, Tyson Tsay, Atharva Mohan, Afnan Shaik, Ruizhe Li, Vasu Sharma, Archana Vaidheeswaran
Prediction markets aggregate information only because traders hold independent beliefs. We show that preference-optimised LLM traders lose that independence: alignment training pulls them toward a shared prior, and the market collapses into a correlated monoculture that prices confidently and wrongly.
FoGen @ ICML 2026
Accepted
Shreya Sanjay Boyane, Sabari Iyyappan Duraipandian, Manju Nagesh, Jerome Francis, Archana Vaidheeswaran, Kevin Zhu
Latent reasoning methods like CODI and COCONUT carry several superimposed candidate traces in hidden space, which makes their reasoning hard to read. We model the latent token sequence as a trajectory in representation space and analyse it as a dynamical system. Latent CoT turns out to have structured, non-random dynamics with two stability classes: CODI behaves as a stable attractor, COCONUT as an unstable expanding system.
FAGEN @ ICML 2026
Accepted
Archana Vaidheeswaran, Param Raval, Rohit Shenoy
LLMs are increasingly proposed as runtime oversight for autonomous agents, but nobody audits the explainer itself. In a black-box audit of an LLM-augmented active inference agent on German energy grid data, 0 of 30 explanations flagged an adversarial 600 MW observation injection, and all three backends rationalised objectively wrong actions 80–95% of the time. Fluent explanation and faithful oversight are different capabilities, and the explainer pattern quietly conflates them.
CAO @ ICLR 2026
Accepted
A framework for measuring how LLM ethical reasoning shifts across linguistic subpopulations. Five models, 50 moral dilemmas, four languages (English, Spanish, Korean, Mandarin) and a seven-pillar rubric: consequentialist bias amplifies outside English and cultural grounding collapses by up to 88%. Language turns out to be a leading indicator of behavioural drift risk in deployed systems.
EACL 2026 SRW
Accepted (41.6% acceptance rate)
We identify fluff injection: replacing a logically necessary reasoning step with plausible-sounding commentary. Our benchmark pairs each problem with a clean chain and a chain where 25–40% of steps are non-inferential filler, same length, same answer. Frontier judges validate the fluffed chains far too often. The verifier we train on it, SmartRM, cuts false positives from 37.4% to 2.7% at 97.3% overall accuracy.
VALED @ WACV 2026
Accepted
Vision-language models are being proposed as scalable tools for judging “substantial similarity” in copyright disputes. Our adversarial “Gaslight Protocol” shows they follow the prompt over the pixels: primed with an authoritative legal accusation, frontier models assign high similarity scores to visually unrelated images. We define a Blindness Index to quantify it, and argue these models are unfit for automated copyright enforcement without human verification.
Preprint
Shubh Laddha, Lucas Changbencharoen, Win Kuptivej, Surya Shringla, Archana Vaidheeswaran, Yash Bhaskar
Tool-retrieval benchmarks tend to use clean, well-formed queries that nobody actually types. HumanMCP is a dataset of human-like queries for measuring whether agents retrieve the right Model Context Protocol tool under realistic phrasing.