Search: “llm”

arXiv:2609.26707v1 Announce Type: cross Abstract: Offline reinforcement learning and off-policy evaluation evaluates dynamic treatment rules based on retrospectively collected data prior to deployment

arXiv:2609.24881v2 Announce Type: new Abstract: In high-stakes decision-making applications of large language models (LLMs), practitioners require not only accurate LLMs but also uncertainty estimates

arXiv:2606.18473v2 Announce Type: replace Abstract: Machine unlearning for large language models (LLMs) aims to remove specified knowledge while preserving the rest of the model's capabilities. Howeve

arXiv:2609.25569v1 Announce Type: new Abstract: Soccer tactics are interactive: an attacking action changes the opponent's defensive problem, and the observed response depends on the multi-agent match

arXiv:2609.25009v1 Announce Type: new Abstract: Numerically equivalent word problems should yield the same canonical answer whether a quantity is written as a decimal, fraction, percentage, number wor

For years, the workday looked remarkably consistent. Employees logged into devices, opened browsers and toggled across a dozen disconnected applications. Users acted as the bridge between apps, copyin

arXiv:2609.26780v1 Announce Type: new Abstract: Long-term conversational memory in multi-party settings requires more than retrieving relevant content from long-term conversations: it must distinguish

arXiv:2609.26131v1 Announce Type: new Abstract: Large language models and vision-language models are increasingly used as high-level planners in robotic systems, using task goals and sensor summaries

arXiv:2503.04332v2 Announce Type: replace-cross Abstract: The tremendous commercial potential of large language models (LLMs) has heightened concerns over their unauthorized use. To address this, we f

There’s an enormous amount of open training data on Hugging Face. What does it take to make it work together 🤗? For Marin’s 535B run, we built on 25T tokens from 152 datasets with licenses permitting

arXiv:2609.26204v1 Announce Type: cross Abstract: When a professional web developer's code fails a test, they do not simply re-read the stack trace. They open the application in a browser, click butto

arXiv:2609.22592v1 Announce Type: cross Abstract: Training agents with reinforcement learning requires a gym, comprising a task, an executable environment in which the task can be attempted, and a ver

arXiv:2609.22124v1 Announce Type: new Abstract: This paper describes the system submitted to the UNLP 2026 Shared Task on Multi-Domain Document Understanding. The challenge required extracting precise

arXiv:2609.22149v1 Announce Type: new Abstract: Multimodal synthetic datasets combine structured attributes with free text, but are often evaluated separately. Such metrics can remain high after tabul

arXiv:2506.12576v3 Announce Type: replace Abstract: Sparse autoencoders (SAEs) can enable inference-time topic steering by modifying latent feature activations, but existing steering methods often fai

arXiv:2609.24264v1 Announce Type: new Abstract: Tool-use agent traces identify messages and API calls, but procedural analyses also need explicit units of action and inspectable links to their evidenc

arXiv:2603.15309v2 Announce Type: replace Abstract: Solving problems through tool use under explicit constraints constitutes a highly challenging yet unavoidable scenario for large language models (LL

Yesterday was Grok 4.7 (pelicans) and MiMo v2.6 Flash/Pro (more pelicans). Today Anthropic released Claude Opus 5.5, and around an hour later OpenAI released GPT-6 Sol and GPT-6 Luna. It's going to ta

Claude Opus 5.5 is the first Opus variant that can automatically fall back from its main model to a reduced‑capability version when executing a small set of specialized tasks associated with frontier

arXiv:2504.14280v2 Announce Type: replace Abstract: As machine learning evolves, domain generalization (DG) and domain adaptation (DA) have become crucial for improving model robustness across diverse

arXiv:2609.22813v1 Announce Type: new Abstract: We present commonsense ranked search (CoRS), a novel path planner that turns an abstract instruction into a route that follows commonsense. While existi

arXiv:2609.22104v1 Announce Type: new Abstract: As automated scientific discovery advances, Large Language Models (LLMs) can now generate research ideas at an unprecedented scale, shifting the bottlen

arXiv:2609.22161v1 Announce Type: cross Abstract: Medical large language models are commonly trained on mixtures of didactic data (e.g., textbooks) and clinical data (e.g., patient records), yet how t

arXiv:2609.23083v1 Announce Type: new Abstract: The distinction between the spirit and letter of the law is a central issue across research and everyday life, and a growing concern for building safe,

arXiv:2609.22223v1 Announce Type: new Abstract: Long-form factuality verification is commonly implemented as a static decompose-search-verify pipeline, with separately prompted modules processing clai

NVIDIA Confidential Computing (CC) secures large‑language‑model inference by providing encrypted virtual machines, GPUs, and NVLink, enabling processing of sensitive data in trusted hardware. Running

arXiv:2404.05880v3 Announce Type: replace Abstract: Jailbreaking attacks can enable Large Language Models (LLMs) to bypass the safeguard and generate harmful content. Existing jailbreaking defense met

arXiv:2609.23573v1 Announce Type: new Abstract: Synthetic essays can help reduce dependence on human-written data in Automated Essay Scoring (AES). However, they often lack realistic errors, limiting

arXiv:2609.22110v1 Announce Type: new Abstract: Background: Large language models (LLMs) can improve healthcare information delivery in low-resource settings but may produce inaccurate or culturally i

arXiv:2609.23726v1 Announce Type: new Abstract: Large language models have shown strong performance across a range of legal tasks, but existing benchmarks rarely evaluate the ability to take and defen

Great paper from Google and colleagues. Trains Text-to-SQL agents using multi-agent RL. (bookmark it) This work proposes DualSQL, which splits Text-to-SQL into two agents, one that links the question

arXiv:2609.23951v1 Announce Type: new Abstract: Expressive speech synthesis has advanced through prosody modeling, yet generating structured poetic speech, such as haiku, remains challenging. Prior wo

Impressive paper showing how much the harness changes a coding agent's results. Harnesses do play a huge role in what you are getting out of the models. GPT-5.5 was run inside both Claude Code and Cod

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

Interesting work on long-horizon research agents. PrimeScientists can decide where a research agent spends its budget. Achieves 10.3% more reward with 50.6% fewer research attempts, under the same bud

Following a successful private preview, we’re thrilled to open DigitalOcean Managed Agents to everyone. Teams can now deploy their preferred agent harness (like OpenCode, Codex CLI) or bring their own

arXiv:2609.23986v1 Announce Type: cross Abstract: Agentic memory is becoming essential for long-horizon AI agents, yet many existing systems rely on autoregressive LLMs to control how memories are org

Been playing with judge models in my eval pipeline, so this landed at the right time. Kev is a small family of Jev-architecture decision models (0.8B, 4B, 9B) built on Qwen3.5, Apache-2.0, with traini

arXiv:2609.23312v1 Announce Type: new Abstract: In environments with large movable obstacles, detour-only navigation can be inefficient or even infeasible, while obstacle interaction requires reasonin

arXiv:2609.22647v1 Announce Type: new Abstract: Visual representations can help lower-primary learners understand Math Word Problems, but generating classroom-usable visuals remains difficult. Existin

arXiv:2502.01562v3 Announce Type: replace Abstract: As the general capabilities of artificial intelligence (AI) agents continue to evolve, their ability to learn to master multiple complex tasks throu

arXiv:2609.24145v1 Announce Type: new Abstract: We present MimicAgent, a prompt-to-trajectory generation framework for learning dynamic quadruped skills. Although reward shaping is extensively used wh

arXiv:2609.24357v1 Announce Type: new Abstract: Cross-domain Named Entity Recognition (CD-NER) aims to transfer the rich knowledge in the source domain to the target domain. Recent studies adopting de

arXiv:2609.24071v1 Announce Type: new Abstract: Chart reasoning agents are increasingly used to extract actionable insights in critical domains, achieving state-of-the-art performance on multiple benc

arXiv:2609.22774v1 Announce Type: new Abstract: This paper presents our system for Track 2 of the NLPCC 2026 Shared Task 10 on citation-level faithfulness in AI-assisted scientific reporting. Given an

arXiv:2609.23959v1 Announce Type: new Abstract: Screening a phone call for fraud needs a trustworthy probability after every caller turn, in milliseconds. Jev-style typed decisions promise exactly tha