Privileged Self-Access Matters for Introspection in AI
arXiv:2508.14802v1 Announce Type: cross Abstract: Whether AI models can introspect is an increasingly important practical question. But there is no ...
Read more
[2410.13098] A Little Human Data Goes A Long Way
[Submitted on 17 Oct 2024 (v1), last revised 20 Aug 2025 (this version, v3)] View a PDF of the paper ...
Read more
A Comprehensive Assessment of Open-Source Large Language Models
[Submitted on 27 Apr 2025 (v1), last revised 19 Aug 2025 (this version, v2)] View a PDF of the paper ...
Read more
Harnessing Frontier Models to Structure Real-World Data at Scale
[Submitted on 2 Feb 2025 (v1), last revised 19 Aug 2025 (this version, v2)] Authors:Cliff Wong, Sam Preston, Qianchu Liu, ...
Read more
[2411.02083] Regress, Don’t Guess — A Regression-like Loss on Number Tokens for Language Models
[Submitted on 4 Nov 2024 (v1), last revised 17 Aug 2025 (this version, v3)] Authors:Jonas Zausinger, Lars Pennig, Anamarija Kozina, ...
Read more
Generalize across Homophily and Heterophily: Hybrid Spectral Graph Pre-Training and Prompt Tuning
arXiv:2508.11328v1 Announce Type: cross Abstract: Graph “pre-training and prompt-tuning” aligns downstream tasks with pre-trained objectives to enable efficient knowledge transfer ...
Read more
ORFuzz: Fuzzing the "Other Side" of LLM Safety — Testing Over-Refusal
arXiv:2508.11222v1 Announce Type: cross Abstract: Large Language Models (LLMs) increasingly exhibit over-refusal – erroneously rejecting benign queries due to overly ...
Read more
Optimizing Rotation Transformation for Quantization for Free
[Submitted on 2 May 2025 (v1), last revised 14 Aug 2025 (this version, v2)] View a PDF of the paper ...
Read more
Unlocking Adaptive Serial-Parallel Decoding by Exploring Intrinsic Parallelism in LLMs
[Submitted on 12 Aug 2025 (v1), last revised 14 Aug 2025 (this version, v2)] View a PDF of the paper ...
Read more
AI Blob! LLM-Driven Recontextualization of Italian Television Archives
arXiv:2508.09535v1 Announce Type: cross Abstract: This paper introduces AI Blob!, an experimental system designed to explore the potential of semantic ...
Read more
