GRPO Training Dynamics for Small Language Models
arXiv:2609.39321v1 Announce Type: cross Abstract: Group Relative Policy Optimization (GRPO) has emerged as a memory-efficient reinforcement fine-tuning (RFT) technique for reasoning-intensive tasks. H
Knowledge catalogue
arXiv:2609.39321v1 Announce Type: cross Abstract: Group Relative Policy Optimization (GRPO) has emerged as a memory-efficient reinforcement fine-tuning (RFT) technique for reasoning-intensive tasks. H
So everyone has been yelling about how I should be using Gufo instead of halogen because it's open source and it's 'just as good or better'. Checking in on their GitHub (GitHub.com/gufo-org/gufo) got
arXiv:2602.11291v3 Announce Type: replace Abstract: World models are becoming central to robotic planning and control by predicting future state transitions. Existing approaches mainly rely on visual,
arXiv:2609.38355v1 Announce Type: new Abstract: We present HalluScoring 2026, a shared task for evaluating hallucination detection and factual verification in Arabic question answering under challengi
arXiv:2609.38760v1 Announce Type: new Abstract: Scalable robotic fleets have become increasingly popular for various applications such as package delivery, warehouse management, and military operation
arXiv:2609.39350v1 Announce Type: cross Abstract: As model sizes continue to scale, distributed training has become inevitable. Automatic parallelization techniques can derive efficient training paral
arXiv:2609.39037v1 Announce Type: cross Abstract: Before a validator can be promoted to a hard gate on a deployment pipeline, it has to be shown that its firing separates outputs that reach users in w
arXiv:2609.38714v1 Announce Type: new Abstract: We describe our winning entry to the Waymo Open Dataset 2D Video Panoptic Segmentation Challenge. The task asks for a semantic class at every pixel of e
arXiv:2609.38291v1 Announce Type: cross Abstract: Large language model (LLM) agents are vulnerable to safety risks such as injected malicious instructions or misleading information, motivating runtime
arXiv:2609.38748v1 Announce Type: new Abstract: Recent joint video-audio generation models have achieved strong semantic correspondence and temporal synchronization. However, applications such as AR/V
arXiv:2609.38332v1 Announce Type: cross Abstract: Test-time scaling improves model performance by allocating additional compute during inference. Using this compute effectively across multiple context
arXiv:2609.39352v1 Announce Type: cross Abstract: LLM agents increasingly rely on reusable Skills for complex, multi-step tasks, creating a critical supply-chain attack surface where poisoned Skill co
arXiv:2506.21734v4 Announce Type: replace Abstract: Reasoning, the process of devising and executing complex goal-oriented action sequences, remains a critical challenge in AI. Current large language
arXiv:2609.38620v1 Announce Type: new Abstract: Neural implicit representations have had a significant impact on scene reconstruction by enabling robots to build continuous, differentiable, and high-f
arXiv:2609.39323v1 Announce Type: cross Abstract: Robot demonstration generation requires a system to identify where an interaction should occur, plan a feasible motion, and execute the required conta
arXiv:2609.39074v1 Announce Type: cross Abstract: Federated learning (FL) on memory-constrained edge devices faces a dilemma: first-order (FO) optimization (i.e., backpropagation) demands substantial
arXiv:2609.38785v1 Announce Type: new Abstract: How accurate must a numerical approximation be within a learning system? Primitive error alone cannot answer this question: errors of the same magnitude
arXiv:2609.39767v1 Announce Type: cross Abstract: The scale and expense of pre-training language models make efficient hyperparameter tuning essential, yet a principled guidance is still missing. In t
arXiv:2609.39336v1 Announce Type: new Abstract: Understanding the fundamental mechanisms of learning is essential for designing systems with strong generalization. Recent studies have shown that incre
arXiv:2609.38917v1 Announce Type: new Abstract: A classifier's conditional accuracy can change while its confidence distribution stays exactly the same. We study the worst-case movement of the reliabi
arXiv:2607.15870v2 Announce Type: replace Abstract: Human label variation in natural language inference is increasingly treated as signal rather than noise, but how much of it formal semantic structur
arXiv:2609.40295v1 Announce Type: new Abstract: Web text makes up the majority of pretraining data and is increasingly AI-generated. After applying FineWeb quality filtering, we find that 27.5% of tok
arXiv:2609.40303v1 Announce Type: new Abstract: Recent autonomous machine learning engineering (MLE) agents have made significant progress on public leaderboards. Often motivated by progress stagnatio
GPT-6 Astra Ultrafast, running on NVIDIA Blackwell GPUs, is available now in the OpenAI API and to eligible ChatGPT Work and Codex users. Accelerated by inference optimizations through OpenAI’s models
arXiv:2609.38257v1 Announce Type: cross Abstract: Code editing requires a model to decide where to make changes, generate the new content, and preserve everything else. We study how masked diffusion l
arXiv:2609.39127v1 Announce Type: new Abstract: Satellite-ground localization estimates the planar position and yaw orientation of a ground camera within a geo-referenced satellite image. Most recent
See how uniopen, a retail platform from Taiwan's Uni-President Enterprises Group, adapted Amazon Nova 2 Lite to its content-moderation policies using supervised fine-tuning in Amazon SageMaker AI and
arXiv:2609.39613v1 Announce Type: new Abstract: Missing data are a fundamental challenge in statistical analysis and machine learning, as the choice of imputation method substantially impacts downstre
arXiv:2507.12133v2 Announce Type: replace Abstract: Device recognition is vital for security in wireless communication systems, particularly for applications like access control. Radio Frequency Finge
arXiv:2609.39550v1 Announce Type: new Abstract: Class-Incremental Learning (CIL) aims to continually learn new classes while preserving prior knowledge. Parameter-efficient fine-tuning with pre-traine
arXiv:1806.11314v3 Announce Type: replace Abstract: Many works have been done on salient object detection using supervised or unsupervised approaches on colour images. Recently, a few studies demonstr
arXiv:2609.39871v1 Announce Type: new Abstract: Hyperspectral remote sensing has advanced across diverse deep learning paradigms, including spectral spatial CNNs, Vision Transformers, Mamba, graph neu
arXiv:2609.40333v1 Announce Type: new Abstract: Self-supervised learning draws inspiration from infant visual development, yet standard training pipelines bear little resemblance to it: images are ind
I have never seen a company more on-brand than OpenAI, one unethical move after another. It's such a full circle moment for OpenAI to use their own AI agents, built on stolen data, to find new and une
I built a free tool that lets your coding agent make videos like this one. You give your agent the link to the repo and say what the video should be about. It writes the words, draws the scenes, and y
I’ve been experimenting with family photographs taken 90–110 years ago, using AI to imagine what the people might look like if the photos were taken today — modern hair, makeup, clothing, body size an
IBM Corp. today announced a self-hosted deployment option for IBM Bob, its agentic software development platform, allowing enterprises to use artificial intelligence to build and modernize application
arXiv:2609.39137v1 Announce Type: new Abstract: Scaling Large Language Models (LLMs) via Mixture-of-Experts (MoE) enables massive parameter growth with nearly constant per-token computation. However,
arXiv:2609.40012v1 Announce Type: new Abstract: Voluntary movements have long been hypothesised to be comprised of discrete primitives called submovements, as a descriptive model of human motor behavi
arXiv:2609.40347v1 Announce Type: new Abstract: We introduce VideoMSN, a Masked Siamese Network framework for efficient self-supervised spatio-temporal representation learning in videos. Instead of re
Learn how to configure secure, multi-environment access to Claude Platform on AWS from a single subscription: cross-account SigV4 for AWS workloads, workspace-scoped API keys for developers, and OIDC
arXiv:2609.39194v1 Announce Type: new Abstract: Wireless split learning (SL) reduces on-device computation by offloading upper layers to a server, yet transmitting high-dimensional intermediate featur
arXiv:2512.19735v4 Announce Type: replace Abstract: Accurately predicting mortality risk in intensive care unit (ICU) patients is critical for clinical decision-making. Large language models (LLMs) ar
arXiv:2609.38282v1 Announce Type: new Abstract: Vision-language models may rewrite anomalous text in images into linguistically plausible expressions, compromising OCR transcription faithfulness. Sequ
arXiv:2609.39215v1 Announce Type: cross Abstract: Time series anomaly detection (TSAD) is increasingly deployed in streaming settings, where data arrive sequentially and may exhibit non-stationarity.
arXiv:2609.38673v1 Announce Type: new Abstract: Model-based offline reinforcement learning (MBORL) improves sample efficiency through model-generated trajectories. However, accumulative model error ca
arXiv:2609.40181v1 Announce Type: new Abstract: We introduce Index-Translate, a multilingual translation model family that combines a shared multilingual foundation with specialized training for gener
arXiv:2609.38362v1 Announce Type: new Abstract: Generative vision-language models (VLMs) such as Qwen-VL and LLaVA achieve strong zero-shot performance on tasks overlapping with their pretraining dist
arXiv:2609.40070v1 Announce Type: cross Abstract: When inference demand exceeds available compute capacity, model providers must decide which requests should be served first. Users have different tole
arXiv:2609.38239v1 Announce Type: cross Abstract: A Technical Report: Operating a large language model (LLM) as a service requires more than inference infrastructure: the provider must also defend aga
arXiv:2609.39406v1 Announce Type: new Abstract: Causal AI is a branch of Artificial Intelligence which helps understand and reason about cause and effect relationships, not just patterns or correlatio
arXiv:2609.39380v1 Announce Type: new Abstract: Reliable infographic generation requires facts, symbols, and visual relations to remain consistent through rendering and revision. Correcting one elemen
arXiv:2504.06209v2 Announce Type: replace Abstract: Predicting future observations plays a central role in machine learning, biology, economics, and many other fields. It lies at the heart of organiza
arXiv:2609.39794v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models have shown strong promise for general-purpose robotic manipulation, yet adapting them to new tasks and domains remai
arXiv:2510.13935v3 Announce Type: replace-cross Abstract: The facts a language model stores are tied to its parameter count, so small models that fit on edge devices fail on expert problems, which nee
arXiv:2606.15911v2 Announce Type: replace Abstract: This paper focuses on automatically generating informative ad descriptions in sponsored search. Unlike ad titles which are usually optimized to attr
During a Swyx podcast interview in October 2026, an OpenAI engineer revealed that the company’s Decision API is built on the GPT‑6 Luna model, augmented with a dedicated decision head. This configurat
arXiv:2609.38625v1 Announce Type: cross Abstract: Concept Bottleneck Models (CBMs) are designed to provide interpretable intermediate representations, yet how such bottlenecks affect robustness remain
Last week we launched DigitalOcean Managed Agents in public preview: a managed runtime, governed access to more than 16,000 tools, serverless inference, and persistent memory and storage, all on one c
We are introducing Clef and Clef-flash, open-source decision models hosted on Workers AI for high-speed classification and agentic workflows. Also launching: a new reinforcement learning platform that