Representational alignment yields generalizable safety in language models
Explorar
Noticias de IA
37834 elementos — filtrados, clasificados y sin duplicados
Not All Preferences Deserve Gradients: Understanding Gradient Utility in Offline Reasonin…
One Editor, Many Edits: A Unified Training-Free Framework for Diverse Video Editing
Knowledge Acquisition During Pre-training? Large Language Models Learn Better With Auxili…
SENTINEL-RL: Offloading Topological Reasoning from LLM Agents in the Security Operations …
A Non-Formulable Theorem: A Fundamental Limit of Finite Syntactic Systems and Its Consequ…
Sequential Beats Joint: On the Interplay between On-Policy Distillation and RLVR
TAP-Path: Task-Adaptive Structural and Token Pruning for Efficient and Trustworthy Pathol…
Influence of Extruded Filament Shape on Buildability in 3D Concrete Printing: A Geometry-…
When Models Edit Too Much: On the Fidelity of Minimal Code Edits
ESPO: Error-Structured Prompt Optimization via Diagnose, Diversify, and Stabilize
RATL: Learning from Retrieved Residuals for Robust Multivariate Time-Series Forecasting
GraFT: A Training-Free Framework for Spatial Reasoning in Multimodal Large Language Model…
Headroom-Drift Replay: A Primitive for Principled Replay Control in GRPO
Investigating the Ability of Large Language Models to Analyze Recipes for Diabetes
FWBC-VLA: Force-Aware Whole-Body Compensation for Contact-Rich Loco-Manipulation
Catalogue Photography as a Cold Start: Toward Deployable Carbide Burr Recognition
Differentiable Interval Bottlenecks for Interpretable Anomaly Detection in Numerical Data
The Blind Spot in 2D Infants' Pose Estimation:Robust Learning from Noisy Annotations
Grammar-Aligned Decoding
ENEAS: Embedding-guided Neural Ensemble for Adaptive Segmentation
IndicSafeEval: Safety Robustness of Large Language Models under Multilingual Persuasive J…
LLaDA-Image: Building Strong Image Generators with Fully Open Training Recipes
Data Market Design through Deep Learning
Out-of-Distribution Generalisation with Sequence Models in Offline Multi-Agent Reinforcem…
Sionna RT: Technical Report
Local Updates, Global Learning (LUGL): Playing Games with non-incremental Learners
Cross-Dataset Transfer and Reliability of Explainable Artificial Intelligence for RhythmF…
Beyond BLEU: A Case for Redefining Sign Language Translation Benchmarks
Witnesses Explain Anomalies