Model releases, architecture work and prompting research on large language models — from frontier-lab announcements to the arXiv papers behind them.
arXiv:2609.24246v1 Announce Type: new Abstract: Large language models (LLMs) have shown remarkable progress in natural language understanding, yet their effectiveness in specialized fields like astro...
arXiv:2609.24259v1 Announce Type: new Abstract: The effectiveness of agent memory ultimately depends on whether the underlying LLM gives each memory in context an appropriate degree of influence over...
The paper investigates an agentic framework for open‑world fake image detection that combines specialist detectors with per‑detector triage, prompting, and conflict‑aware evidence arbitration. Experiments across six configurations and three multimodal large language model backbones reveal that naive detector fusion yields high false‑positive rates, while triage and prompting consistently filter unreliable evidence. The most significant improvement comes from the reasoning component: a stronger judge markedly outperforms a weaker one, especially under distribution shift, and overall manipulation recall is nearly saturated, highlighting that the key challenge lies in calibrating trust and arbitrating conflicting forensic evidence rather than detecting manipulations themselves.
arXiv:2609.24380v1 Announce Type: new Abstract: RLVR has substantially improved the reasoning capabilities of LLMs. However, existing methods typically parameterize temporal progression in the Markov...
arXiv:2609.24417v1 Announce Type: new Abstract: Despite advances in long-context inference, large language models (LLMs) remain fundamentally limited by the key-value (KV) caching mechanisms that are...
arXiv:2609.24424v1 Announce Type: new Abstract: Monocular RGB-based hand pose estimation has emerged as a critical research frontier in computer vision. The local hand pose estimation methods predict...
arXiv:2609.24485v1 Announce Type: new Abstract: Visual token pruning is a promising approach to reducing the inference cost of large vision-language models (LVLMs), yet aggressive token reduction oft...
arXiv:2609.24487v1 Announce Type: new Abstract: In this work, we present a method for shape reconstruction and tracking from video via agentic analysis-by-synthesis. Unlike prior methods which first...
arXiv:2609.24538v1 Announce Type: new Abstract: Functional annotation of newly sequenced proteins remains a bottleneck in molecular biology: the number of sequences in public repositories grows far f...
arXiv:2609.24559v1 Announce Type: new Abstract: We present $t_0$, a family of open-weights foundation models for forecasting with multivariate context. We release its first two members: $\texttt{t0-a...
arXiv:2609.24646v1 Announce Type: new Abstract: On-policy self-distillation fine-tuning (SDFT) learns new skills from demonstrations while reducing forgetting, but it always distils toward the full d...
arXiv:2609.24698v1 Announce Type: new Abstract: Repeated execution of the target model during autoregressive decoding is a major source of LLM inference latency. Unlike linear speculation, which foll...
arXiv:2609.24746v1 Announce Type: new Abstract: Existing approaches to solving differential equations, such as symbolic regression, physics informed neural networks, and neural operators, typically f...
PrismGPT is a Vision‑Language Model that generates structured, region‑aware photo‑editing plans from a single image, without relying on commercial black‑box tools. It learns to diagnose aesthetic issues globally and locally while predicting precise editing parameters, using proxy‑guided learning with operation decomposition and region‑aware aesthetic ranking to bootstrap the model. A competence‑based dynamic scheduler shifts training focus from proxy tasks to the main editing task as skills improve, and all reasoning traces for fine‑tuning are self‑synthesized by the model itself. Experiments on MIT‑Adobe FiveK and a new professionally retouched benchmark, SPIRE, show PrismGPT achieves state‑of‑the‑art results using only about 6% of the training data required by previous methods.
arXiv:2609.24890v1 Announce Type: cross Abstract: Evaluation of Computer-Use Agents (CUAs) is often limited to the final deliverables they create (at the end of hundreds of steps) and assessed with f...
arXiv:2108.11204v4 Announce Type: replace-cross Abstract: Humans excel in solving complex reasoning tasks through a mental process of moving from one idea to a related one. Inspired by this, we propo...
arXiv:2410.23912v3 Announce Type: replace-cross Abstract: The reasoning abilities of large language models (LLMs) have improved with chain-of-thought (CoT) prompting, allowing models to solve complex...
arXiv:2604.17614v2 Announce Type: replace-cross Abstract: Skills are a natural unit for describing what a language model can do and how its behavior can be changed. However, existing characterization...
arXiv:2505.16227v4 Announce Type: replace-cross Abstract: Personalizing jargon detection and explanation is essential for making technical documents accessible to readers with diverse disciplinary ba...
Lingshu is a medical‑specialized multimodal large language model that addresses key limitations of existing medical MLLMs, such as narrow knowledge coverage, hallucinations, and weak reasoning. The authors curate a comprehensive dataset combining medical imaging, texts, and general‑domain data, then train Lingshu in multiple stages to embed medical expertise and improve task performance. They also introduce MedEvalKit, a unified evaluation framework, and demonstrate that Lingshu outperforms current open‑source multimodal models on multimodal QA, text‑based QA, and medical report generation.