Skip to main content
Search

Indian Institute of Technology Patna

4 related articles

Subscribe to updates
  1. arXiv

    MEA uses reward-driven multi-agent optimization to raise explanation faithfulness by up to 34%

    The work presents MEA, a multi-agent framework in which a Proposer agent selects and configures explanation tools based on the question and modality while an Actor agent is optimized end-to-end against faithfulness to turn outputs into natural-language explanations; the authors also introduce question types spanning feature attribution, counterfactual reasoning, and spurious feature detection, each paired with a perturbation-based faithfulness metric, find that frontier LLMs systematically produce unfaithful explanations, and report that MEA outperforms post hoc explainers, agentic, and closed-source baselines across six datasets, with faithfulness gains of +28% (tabular), +21% (text), and +34% (vision) over the untrained backbone.

  2. A hybrid Random Committee and Multilayer Perceptron Regressor model predicts particle Froude number in auto-washout drainage systems with sedimented beds and identifies volumetric sediment concentration as the most sensitive variable

    Using a Multilayer Perceptron Regressor (MLPR) as the base model and a Random Committee hybrid (RC-MLPR), the study predicted the particle Froude number (PFr) in auto-washout drainage systems from five heterogeneous datasets collected from existing literature covering a wide range of hydraulic and sediment conditions, evaluated the models with several performance measures including the agreement index, found that RC-MLPR outperformed other proposed ML models, state-of-the-art ML models, and existing empirical equations, and reported from sensitivity analysis that volumetric sediment concentration (Csed) is the most sensitive variable for PFr prediction by the hybrid RC-MLPR model.

  3. ChemPhysChem

    Gemini surfactant chain length and spacer tune micelle structure, while NaSal grows micelles from 2.6 nm to 15.8 nm and lengthens MC540 fluorescence lifetime

    Using tensiometry, dynamic light scattering, small-angle neutron scattering, and steady-state and time-resolved fluorescence at 30 °C, this study examined four Gemini surfactants (8-4-8, 10-4-10, 12-4-12, 10-8-10) at 100 mM with and without 100 mM NaBr or NaSal, and found that longer alkyl chains lower the CMC (about 55.4 mM to about 1.2 mM) and enlarge micelles with higher aggregation numbers (13 to 39), that increasing the spacer from 4 to 8 carbons shrinks micelles (D_h about 2.6 nm to about 1.1 nm), that NaBr swells micelles by electrostatic screening (D_h about 4.8 nm) while NaSal elongates them markedly (D_h about 15.

  4. arXiv

    MemGuard-Alpha audit finds MIA contamination scores drop to 0.487–0.537 discrimination at a fixed date, and no filtering variant beats the unfiltered ensemble under symmetric transaction costs

    The authors introduce MemGuard-Alpha, comprising a composite contamination score combining five membership inference attack (MIA) methods with a temporal proximity feature, plus Cross-Model Memorization Disagreement exploiting variation in training cutoffs across models, and audit both across seven LLMs (124M–7B), 50 S&P 100 constituents, 42,800 prompts and 299,600 prompt-model MIA scores spanning 2019–2024, yielding three negative results: the temporal proximity feature recovers the in-sample label perfectly (ROC-AUC 1.000) because it is a monotone transform of the defining variable, the discriminative power of MIA scores is largely attributable to scale differences between models (raw scores reach AUC up to 0.99 at a fixed date but fall to 0.487–0.

You’re all caught up