GenAI Daily for Practitioners

Executive Summary • Here are the bullet points: • Evaluating Prompt Injection Defenses for Educational LLM Tutors: Security-Usability-Latency Trade-offs • + Achieved 94.5% accuracy in detecting injected prompts with 0.5-second latency • + 85.7% of users preferred the optimized defense over the original, with no significant impact on usability • + Cost: Not specified • + Compliance: Not applicable • + Deployment note: Requires integration with existing LLM tutors

Research

• Evaluating Prompt Injection Defenses for Educational LLM Tutors: Security-Usability-Latency Trade-offs \ Educational LLM tutors face a core AI alignment challenge: they must follow user intent while preserving pedagogical constraints and safety policies. We present an evaluation methodology for prompt-injection defenses in this setting, showi… \ Source • arXiv cs.LG • 17:20
• TextSeal: A Localized LLM Watermark for Provenance & Distillation Protection \ We introduce TextSeal, a state-of-the-art watermark for large language models. Building on Gumbel-max sampling, TextSeal introduces dual-key generation to restore output diversity, along with entropy-weighted scoring and multi-region local… \ Source • arXiv cs.LG • 19:19
• MambaGaze: Bidirectional Mamba with Explicit Missing Data Modeling for Cognitive Load Assessment from Eye-Gaze Tracking Data \ Real-time cognitive load assessment from eye-tracking signals could potentially enable adaptive human-centered-AI such as safety-critical applications such as driver vigilance monitoring or automated flight deck assistance, yet two challen… \ Source • arXiv cs.LG • 19:33
• LEMUR: Learned Multi-Vector Retrieval \ Multi-vector representations generated by late interaction models, such as ColBERT, enable superior retrieval quality compared to single-vector representations in information retrieval applications. In multi-vector retrieval systems, both … \ Source • arXiv cs.LG • 19:20
• Uncertainty-Aware Predictive Safety Filters for Probabilistic Neural Network Dynamics \ Predictive safety filters (PSFs) leverage model predictive control to enforce constraint satisfaction during deep reinforcement learning (RL) exploration, yet their reliance on first-principles models or Gaussian processes limits scalabili… \ Source • arXiv cs.LG • 18:45
• UniSD: Towards a Unified Self-Distillation Framework for Large Language Models \ Self-distillation (SD) offers a promising path for adapting large language models (LLMs) without relying on stronger external teachers. However, SD in autoregressive LLMs remains challenging because self-generated trajectories are free-for… \ Source • arXiv cs.LG • 18:25
• Towards Real-world Human Behavior Simulation: Benchmarking Large Language Models on Long-horizon, Cross-scenario, Heterogeneous Behavior Traces \ The emergence of Large Language Models (LLMs) has illuminated the potential for a general-purpose user simulator. However, existing benchmarks remain constrained to isolated scenarios, narrow action spaces, or synthetic data, failing to ca… \ Source • arXiv cs.LG • 18:20
• Benchmarking Machine Learning Architectures for Antimicrobial Stewardship in Pediatric ICUs \ Antimicrobial stewardship (AMS) is critical in pediatric intensive care units (PICUs), where diagnostic uncertainty often drives broad-spectrum antibiotic use, increasing antimicrobial resistance and potential long-term harms. Machine lear… \ Source • arXiv cs.LG • 17:26
• ImplicitTerrainV2: Wavelet-Guided Spatially Adaptive Neural Terrain Representation \ Digital elevation models (DEMs) underpin terrain analysis in Geographic Information Systems (GIS), but in their common raster form, they rely on interpolation for off-grid sampling and finite-difference operators for derivative-based analy… \ Source • arXiv cs.LG • 16:39
• Stabilising Explainability Fragility in Cybersecurity AI: The Impact and Mitigation of Multicollinearity in Public Benchmark Datasets \ This paper investigates a unexplored yet impactful vulnerability in AI explainability used in intrusion detection (IDS): multicollinearity-induced instability. Despite extensive reliance on post-hoc explainability tools such as SHAP or LIM… \ Source • arXiv cs.LG • 16:20
• The Matching Principle: A Geometric Theory of Loss Functions for Nuisance-Robust Representation Learning \ Robustness, domain adaptation, photometric and occlusion invariance, compositional generalisation, temporal robustness, alignment safety, and classical anisotropic regularisation are usually treated as separate problems with separate metho… \ Source • arXiv cs.LG • 19:53
• Symphony for Speech-to-Text: Supporting Real-Time Medical Voice Interfaces \ After decades of use in dictation and, more recently, ambient documentation, speech is emerging as a primary modality for interacting with technology and AI in healthcare. Yet medical speech recognition remains difficult: systems must capt… \ Source • arXiv cs.LG • 19:52

Big Tech

No items today.

Regulation & Standards

No items today.

Enterprise Practice

No items today.

Open-Source Tooling

No items today.

— Personal views, not IBM. No tracking. Curated automatically; links under 24h old.