arXiv:2607.09709v2 Announce Type: replace Abstract: Post-training a code generator against a learned judge can optimize proxy features that raise the…
Category: cs.AI updates on arXiv.org
FADE: Mitigating Hallucinations by Reducing Language-Prior Dominance in Large Vision-Language Models
arXiv:2606.29431v5 Announce Type: replace Abstract: Despite the impressive capabilities of Large Vision-Language Models (LVLMs), they remain susceptible…
Shared Selective Persistent Memory for Agentic LLM Systems
arXiv:2607.09493v2 Announce Type: replace Abstract: Agentic LLM systems that generate code through multi-turn tool use face a fundamental context problem:…
RegNetAgents: A Multi-Agent Framework for Cross-Network Regulatory Driver Identification in Cancer Genomics
arXiv:2607.14097v2 Announce Type: replace Abstract: We introduce RegNetAgents, an AI-oriented multi-agent framework for structured, query-driven…
ANCHOR: An External LLM-Driven Supervisory Module Facilitating Healthy Evolution in Self-Evolving Systems
arXiv:2606.06114v3 Announce Type: replace Abstract: Self-evolving agents improve through continual self-play and self-generated learning signals, but…
Measuring Curriculum Alignment across Topical Coverage, Competency, and Cognitive Depth: A Longitudinal Framework Applied to CS2013 and CS2023
arXiv:2606.19469v2 Announce Type: replace Abstract: Undergraduate computer science is governed by international curricular guidelines revised about once a…
Shielded Analysis: Certification and Characterization of Defensibility in Systems under Adversarial Interaction
arXiv:2606.13621v2 Announce Type: replace Abstract: Formal safety analysis determines whether a system admits a safe defense; adaptive evaluation…
Superficial Beliefs in LLM Decision-Making
arXiv:2606.11016v2 Announce Type: replace Abstract: We ask whether large language models (LLMs) merely imitate rationales when choosing between two…
Rhythm of the Deep: Two-tier acoustic organization of sperm-whale codas from click waveforms to second-order sequence dependence
arXiv:2606.16084v3 Announce Type: replace Abstract: Sperm-whale codas are conventionally characterized by click count and inter-click intervals (ICIs),…
How Clinicians Think and What AI Can Learn From It
arXiv:2601.12547v2 Announce Type: replace Abstract: Clinical artificial intelligence increasingly builds high-dimensional representations of patients, yet…