arXiv AI
Aug 6
Gradient Immunity: Null-Space Resistance to Malicious Fine-Tuning
arXiv:2608. 05045v1 Announce Type: cross Abstract: Released aligned large language models remain vulnerable to malicious downstream finetuning.
AI safety and alignment
Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.
arXiv Machine Learning
Aug 6
Toward Skill-Native LLMs: Skill Entropy for Benchmarking and Training Long-Horizon Reasoning
arXiv:2608. 05139v1 Announce Type: cross Abstract: Long-horizon reasoning in recent LLMs demands that the model switch between distinct skills inside a reasoning chain, such as first doing a math derivation, then using the result to plan a schedule.
arXiv AI
Aug 6
SafeCommit: Certifying When Memory-Grounded Agents May Safely Act
arXiv:2608. 04289v1 Announce Type: new Abstract: Long-horizon agents increasingly use persistent memory and tools to take actions with external side effects.
arXiv AI
Aug 6
Modality Agreement- and Conflict-Aware Prototype Hypergraph Learning for Multimodal Intent Understanding
arXiv:2608. 04054v1 Announce Type: cross Abstract: Multimodal intent recognition requires understanding not only what textual, acoustic, and visual signals share, but also how they disagree.
arXiv AI
Aug 6
Approximate Multi-Objective Search Under Rulebooks
arXiv:2608. 04398v1 Announce Type: cross Abstract: Robotic planning often involves multiple objectives with complex priority relationships, such as safety, efficiency, and regulatory compliance.
arXiv AI
Aug 6
Assessing and Explaining the Persuadability of Large Language Models as Legal Decision Tools
arXiv:2604. 26233v3 Announce Type: replace Abstract: As Large Language Models (LLMs) are proposed as legal decision assistants, and even first-instance decision-makers, across a range of judicial and administrative contexts, it becomes essential to explore how they answer legal questions, and in particular the factors that lead them to decide difficult questions.
arXiv AI
Aug 6
Robust and Efficient Motion Reasoning for Privacy-Aware Classroom Incident Recognition
arXiv:2608. 05115v1 Announce Type: cross Abstract: Can computer vision help make classrooms safer?
arXiv AI
Aug 6
Dynamic Jailbreaking Attack
arXiv:2510. 02422v4 Announce Type: replace-cross Abstract: Existing gradient-based jailbreak attacks typically optimize a fixed-length adversarial suffix toward a predefined target response with a static optimization strategy.
arXiv Machine Learning
Aug 6
Contrastive Diffusion Alignment: Learning Structured Latents for Controllable Generation
arXiv:2510. 14190v3 Announce Type: replace Abstract: Diffusion models excel at generation, but their latent spaces are high dimensional and not explicitly organized for interpretation or control.
arXiv Machine Learning
Aug 6
Multicalibration Yields Better Matchings
arXiv:2511. 11413v2 Announce Type: replace Abstract: Consider the problem of finding the best matching in a weighted graph where we only have access to predictions of the actual stochastic weights, based on an underlying context.
arXiv AI
Aug 6
CARGO-VL: Counterfactual Arbitration with Risk-Constrained Group Optimization for Vision-Language Models
arXiv:2608. 04509v1 Announce Type: new Abstract: Vision-language systems combine images with retrieved text, but these sources can disagree or jointly fail to support an answer.
arXiv AI
Aug 6
Neural Diversity Regularizes Hallucinations in Language Models
arXiv:2510. 20690v3 Announce Type: replace-cross Abstract: Language models continue to hallucinate despite increases in parameters, compute, and data.
arXiv AI
Aug 6
Long-term Measurements: Towards a Longitudinal Understanding of Human-AI Interactions
arXiv:2608. 02491v2 Announce Type: replace Abstract: Language models have taken on the role of a very new type of technology, by virtue of their "human-ness" and rapid integration into users' daily lives.
arXiv AI
Aug 6
Chain-of-Thought Monitoring Can Be Unreliable in Implicit-Influence Settings
arXiv:2608. 04735v1 Announce Type: new Abstract: Chain-of-thought (CoT) monitoring is increasingly treated as an important safety layer for frontier reasoning models.
arXiv AI
Aug 6
Calibrating Transformer Attention via Task-Space Sensitivity Feedback
arXiv:2512. 20661v2 Announce Type: replace Abstract: Transformer-based pre-trained language models (PLMs) excel in text classification but suffer from attention dilution and attention sink effects, forcing models to over-focus on task-irrelevant tokens.
arXiv Machine Learning
Aug 6
The Neural Echo: A Signal Processing Perspective for Understanding Neural Networks
arXiv:2608. 04864v1 Announce Type: cross Abstract: We introduce the neural echo as a tool for understanding the behavior of neural networks.
arXiv AI
Aug 6
AudioScape-TTA: A Structured Soundscape Benchmark for Fine-Grained Text-to-Audio Evaluation
arXiv:2608. 04479v1 Announce Type: cross Abstract: Text-to-audio (TTA) generation has recently achieved remarkable progress in synthesizing realistic audio from natural language descriptions.
arXiv AI
Aug 6
NSF-HRPT: Neural Semantic Field meets Hierarchical Risk Perception Tree for Safety-Critical Scenario Assessment
arXiv:2608. 04776v1 Announce Type: new Abstract: The ability to accurately assess and anticipate risks in safety-critical scenarios is crucial for autonomous driving systems.
arXiv AI
Aug 6
Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation
arXiv:2608. 04794v1 Announce Type: new Abstract: Self-distillation (SD) has emerged as a compute-efficient alternative to reinforcement learning with verifiable rewards: a self-teacher, conditioned on privileged information (PI) about the answer such as a reference solution, supplies dense per-token supervision to a student that never sees it.
arXiv AI
Aug 6
Easy to Complete, Hard to Choose: Investigating LLM Performance on the ProverbIT Benchmark
arXiv:2608. 04670v1 Announce Type: cross Abstract: Large Language Models (LLMs) have transformed computational linguistics and achieved remarkable performance across numerous natural language processing tasks, yet significant gaps persist in understanding how these systems process culturally embedded linguistic expressions.