Model releases, architecture work and prompting research on large language models — from frontier-lab announcements to the arXiv papers behind them.
arXiv:2608. 07905v1 Announce Type: new Abstract: Embodied agents using LLM-based planners often struggle with physical hallucinations, poor generalization to long-horizon tasks, and lack of environmental awareness.
arXiv:2608. 07925v1 Announce Type: new Abstract: EDA scripting with tool-specific, often undocumented APIs remains a long-tail bottleneck that existing LLMs fail to address.
arXiv:2608. 08189v1 Announce Type: new Abstract: LLM-driven program discovery relies on rapid evaluator feedback, but many scientific and engineering tasks require high-fidelity simulations, hardware execution, or physical experiments, making each evaluation expensive.
arXiv:2608. 07481v1 Announce Type: cross Abstract: This paper investigates whether one large language model can approximate the humor preferences of another in a controlled Cards Against Humanity-style task.
arXiv:2608. 08026v1 Announce Type: new Abstract: We investigate how social authority (SA) signals interact with severity-based prioritization in large language models, operationalizing each axis as a model-elicited baseline -- the triage hierarchy and the SA hierarchy.
arXiv:2608. 08037v1 Announce Type: new Abstract: LLM-based agent frameworks now act as personal assistants for multi-step tasks.
arXiv:2608. 08176v1 Announce Type: new Abstract: On-policy self-distillation (OPSD) improves the reasoning abilities of LLMs by internalizing privileged context into model parameters through self-distillation.
arXiv:2608. 08212v1 Announce Type: new Abstract: In-context learning (ICL) can induce emergent misalignment (EM), where narrow misaligned examples alter answers to unrelated questions.
arXiv:2608. 08210v1 Announce Type: new Abstract: Collaborative dialogue can end with apparent agreement while participants still differ on goals, assumptions, or execution plans, creating an \textbf{illusion of alignment (IoA)}.
arXiv:2608. 08640v1 Announce Type: new Abstract: Large language model agents increasingly rely on reusable skills to extend their capabilities beyond parametric knowl- edge.
arXiv:2608. 08264v1 Announce Type: new Abstract: Large language model agents are becoming operational interfaces to files, memories, registries, and external tools.
arXiv:2608. 08382v1 Announce Type: new Abstract: As LLM inference shifts to multi-tenant GPU clusters, co-batching improves throughput but obscures per-tenant usage and limits control.
arXiv:2608. 08303v1 Announce Type: new Abstract: Agentic skills improve large language model (LLM) agents by encoding reusable procedures for complex tasks.
arXiv:2608. 07631v1 Announce Type: cross Abstract: LLM-based full-duplex voice services allow users to speak while the assistant is responding.
arXiv:2608. 08453v1 Announce Type: new Abstract: Under the current standard, Agent Skills are SKILL.
arXiv:2608. 08506v1 Announce Type: new Abstract: Training-free low-rank compression frameworks have been gaining prominence for LLM compression given their effectiveness in reducing model parameter count while maintaining task-level accuracy.
arXiv:2608. 08677v1 Announce Type: new Abstract: Skill evolution improves agent skills through feedback over time, with failed trajectories often providing informative signals by revealing incomplete or misleading behaviors.
arXiv:2608. 08691v1 Announce Type: new Abstract: Residential virtual power plants (VPPs) can provide grid flexibility by shifting household demand, but physical flexibility becomes dependable capacity only when residents authorize a plan and the promised response is delivered.
arXiv:2512. 20968v2 Announce Type: replace-cross Abstract: Distributed attention is essential for scaling large language models (LLMs) to long contexts, yet existing methods either have limited parallelism or incur high communication costs.
arXiv:2607. 23191v3 Announce Type: replace Abstract: Fine-tuned code LLMs are routinely conditioned on a design-intent specification, but the correctness axis of such a signal -- a wrong intent rather than an absent one -- has not been tested, and the benefit of conditioning is usually scored with the same detector that defines the signal.