Kargi Chauhan

Kargi Chauhan

AI scientist systems are beginning to automate the production, evaluation, and iteration of scientific hypotheses. Their promise is speed; their risk is that speed also scales errors embedded in the scientific record. We argue that a near-term risk is corpus failure: AI scientist systems are trained on and grounded in a literature that over-represents positive results and under-represents null findings. We formalise this distortion as the null result gap, estimate it across three domains (drug discovery ~0.60, psychology ~0.56, cancer biology ~0.35), and introduce an amplification index for reasoning about how retrieval, generation, and automated evaluation can compound the raw gap. Using first-order estimates, we argue that a standard three-stage pipeline can amplify corpus distortion by a factor of 2.18x, with the conclusion unchanged under more conservative multipliers. We identify four governance failure modes: confident rediscovery, ghost evidence accumulation, replication laundering, and confidence miscalibration. We then propose three interventions: null-result databases as training infrastructure, retraction-aware evaluation metrics, and mandatory training corpus disclosure. The central takeaway is that AI scientists will not only accelerate science. Without governance, they will accelerate science's blind spots before they accelerate its discoveries.

Kargi Chauhan, Pratibha Revankar

LLM agents often place sensitive credentials in the same context window as untrusted retrieved content, creating a direct path for indirect prompt injection to induce credential exfiltration. We study this failure mode through three complementary defenses. First, we ask whether activation probes can detect credential access before output tokens are emitted. Second, we construct honeytokens from format-specific character models and calibrate detection with split conformal prediction. Third, we treat multi-turn exfiltration as a cumulative information-flow problem and track an estimated leakage budget across conversation turns. In controlled experiments on open-weight models, activation features separate benign and credential-seeking prompts with high accuracy, including under held-out encoding transformations. In a small synthetic multi-turn suite, cumulative accounting detects attacks that per-turn detectors miss. These results are preliminary: the multi-turn benchmark is in-house and small, the activation method requires white-box access, and the information estimator provides a practical signal rather than a formal upper bound. Still, the results suggest that credential-exfiltration defenses should combine pre-output monitoring, calibrated canary detection, and temporal leakage accounting rather than relying only on text-level output filters.

Kargi Chauhan

AI scientist systems increasingly choose biological foundation models before they choose experiments. In protein pipelines, this creates a concrete engineering and scientific question: when is the cost of structural inference worth paying over a cheaper sequence-only model? We introduce the information bonus (IB), a task-level metric that measures the linearly accessible advantage of frozen single-sequence AlphaFold2 Evoformer representations over frozen ESM-2 embeddings under protein-level cross-validation. Across binding affinity regression (PDBbind, n=5,680), conformational flexibility (ATLAS molecular dynamics, 268 proteins), and allosteric-site classification (AlloSigDB, n=9,925 residues), IB is sharply mechanism-dependent. ESM-2 dominates binding affinity (IB=-0.141; Pearson r=0.449 vs. 0.307) and binary flexibility (IB=-0.060; AUROC 0.824 vs. 0.764; p=0.0017). AF2 single representations give the only above-chance allostery predictions (IB=+0.064; AUROC 0.548 vs. 0.485), revealing long-range geometric signal not recovered from sequence alone. We also identify a residue-level leakage artifact: naive residue splits inflate RMSF performance by 27-39% depending on the representation, enough to reverse representation rankings. These results turn representation selection into a measurable decision for AI-for-science systems.

Kargi Chauhan, Sadiba Nusrat Nur

With the rapid growth of large language models for code generation, distinguishing between human-written and AI-generated code has become increasingly critical for academic integrity, hiring evaluations, and software security. We present our system for SemEval-2026 Task 13: Multilingual Machine-Generated Code Detection, participating in Subtask A (binary detection) and Subtask B (multi-class attribution across 10 LLM families). For Subtask A, we fine-tune UniXcoder-base with a multi-view training framework that promotes generator-invariant representations. The framework combines domain-specific structural prefixes, delexicalization with symmetric KL consistency loss, token dropout, and mixed-content augmentation. Our system achieves 0.993 macro F1 on validation and 0.845 macro F1 on the test set, which spans unseen languages and domains. For Subtask B, we show that severe class imbalance (88.4% human code, 221:1 majority-to-minority ratio) causes catastrophic minority-class failure under standard fine-tuning, with macro F1 collapsing to 0.086 despite 88.4% accuracy. A class-weighted extension trained for 3 epochs recovers macro F1 to 0.345 (+301% relative), confirming that multi-class attribution requires imbalance-aware training strategies.

Kargi Chauhan, Mahalakshmi Venkateswarlu

Large Language Models (LLMs) are increasingly deployed for personalized product recommendations, with practitioners commonly assuming that longer user purchase histories lead to better predictions. We challenge this assumption through a systematic benchmark of four state of the art LLMs GPT-4o-mini, DeepSeek-V3, Qwen2.5-72B, and Gemini 2.5 Flash across context lengths ranging from 5 to 50 items using the REGEN dataset. Surprisingly, our experiments with 50 users in a within subject design reveal no significant quality improvement with increased context length. Quality scores remain flat across all conditions (0.17--0.23). Our findings have significant practical implications: practitioners can reduce inference costs by approximately 88\% by using context (5--10 items) instead of longer histories (50 items), without sacrificing recommendation quality. We also analyze latency patterns across providers and find model specific behaviors that inform deployment decisions. This work challenges the existing ``more context is better'' paradigm and provides actionable guidelines for cost effective LLM based recommendation systems.

Kargi Chauhan, Leilani H. Gilpin

Modern diffusion models generate realistic traffic simulations but systematically violate physical constraints. In a large-scale evaluation of SceneDiffuser++, a state-of-the-art traffic simulator, we find that 50% of generated trajectories violate basic physical laws - vehicles collide, drive off roads, and spawn inside buildings. This reveals a fundamental limitation: current models treat physical validity as an emergent property rather than an architectural requirement. We propose Validity-First Spatial Intelligence (VFSI), which enforces constraints through energy-based guidance during diffusion sampling, without model retraining. By incorporating collision avoidance and kinematic constraints as energy functions, we guide the denoising process toward physically valid trajectories. Across 200 urban scenarios from the Waymo Open Motion Dataset, VFSI reduces collision rates by 67% (24.6% to 8.1%) and improves overall validity by 87% (50.3% to 94.2%), while simultaneously improving realism metrics (ADE: 1.34m to 1.21m). Our model-agnostic approach demonstrates that explicit constraint enforcement during inference is both necessary and sufficient for physically valid traffic simulation.