Saisab Sadhu

Large language models (LLMs) exhibit a curious failure mode in retrieval-augmented generation: they may internally "disagree" with retrieved context even while outwardly appearing to follow it. We identify a mechanistic law (p < 10^-42) predicting internal epistemic conflict in 70B-scale models by examining logit-level interactions between parametric knowledge and retrieved context. Our analysis uncovers the Alignment Paradox: instruction-tuning decouples a model's internal epistemic tension from its surface-level textual behavior, causing standard behavioral metrics to systematically miss genuine conflict. We further develop a 25ms "Mechanistic Auditor" that identifies latent sycophancy with 76% F1, achieving a 600× speedup over state-of-the-art probing methods and significantly outperforming model self-reporting.

We present DARE (Dialectical Adversarial and Evidence-Aware RAG), a framework that resolves factual conflicts in retrieval-augmented generation through a structured cross-examination process. Rather than weighting sources by static relevance scores, DARE subjects candidate passages to adversarial challenges and infers source reliability from logical resilience — how well each source withstands targeted counterfactual pressure. This dynamic credibility assessment mechanism achieves state-of-the-art gains of 77% on FaithEval and 28% on RAMDocs. The framework is model-agnostic and operates entirely at inference time, requiring no additional fine-tuning.

Generating high-quality financial analysis from earnings call transcripts requires synthesizing heterogeneous signals — management tone, analyst pushback, forward guidance, and market context — into coherent, persuasive reports. We design a hierarchical multi-agent framework modelling investment committee debates: specialist agents independently analyse distinct aspects of the call, a moderator orchestrates structured argumentation, and a synthesis agent produces the final report. Our system achieved a 68.75% win rate over cooperative baselines, ranking first globally on the official "Win Rate vs. Analyst Report" metric — preferred over professional human analysts.

Abstractive summarization of ultra-long legal documents presents unique challenges: legal text is structured by rhetorical roles (facts, arguments, holdings, orders), and naive chunking destroys cross-section coherence. We propose a rhetorically-informed chunking pipeline that segments documents along argumentative boundaries before abstractive generation. Through systematic analysis we identify and characterize the "Coherence Gap" — a fundamental trade-off between local phrase-level accuracy and global narrative coherence in legal summarization — and propose mitigation strategies through structure-aware segmentation.

AI tutoring systems are prone to confident errors, sycophantic validation, and pedagogically unsound explanations. We introduce Hierarchical Pedagogical Oversight (HPO), a multi-agent adversarial framework for reliable AI tutoring. HPO employs a hierarchy of specialised agents — a tutor, a challenger, and an overseer — where the challenger actively probes for errors and the overseer arbitrates. An 8B-parameter model structured via HPO outperforms GPT-4o by 3.3% Macro F1 on MRBench, demonstrating that adversarial architectural design can overcome raw model scale for pedagogical reliability.