model releases

Sorry for the long wait, everyone! As I said, Qwen is going to keep open-sourcing! ⚡ Meet Qwen3.6-35B-A3B：Now Open-Source！🚀🚀 A sparse MoE model, 35B total params, 3B active. Apache 2.0 license. 🔥 Agen

arXiv:2604.13847v1 Announce Type: new Abstract: While sparse attention mitigates the computational bottleneck of long-context LLM training, its distributed training process exhibits extreme heterogene

arXiv:2604.14144v1 Announce Type: cross Abstract: Spatial reasoning over three-dimensional scenes is a core capability for embodied intelligence, yet continuous model improvement remains bottlenecked

arXiv:2604.13123v1 Announce Type: new Abstract: Grokking -- delayed generalisation long after memorisation -- lacks a predictive mechanistic explanation. We identify the normalised spectral entropy il

arXiv:2604.01034v2 Announce Type: replace Abstract: We propose a Stein variational distributionally robust controller for nonlinear dynamical systems with latent parametric uncertainty. The method is

arXiv:2604.14017v1 Announce Type: cross Abstract: Under interpolation-type assumptions such as the strong growth condition, stochastic optimization methods can attain convergence rates comparable to f

arXiv:2604.13620v1 Announce Type: new Abstract: Managing natural dialogue timing is a significant challenge for voice-based chatbots. Most current systems usually rely on simple silence detection, whi

arXiv:2604.13192v1 Announce Type: cross Abstract: Robust control barrier functions (CBFs) provide a principled mechanism for smooth safety enforcement under worst-case disturbances. However, existing

arXiv:2604.13538v1 Announce Type: new Abstract: Using responses generated by high-performing large language models (LLMs) for instruction tuning has become a widely adopted approach. However, the exis

arXiv:2604.13125v1 Announce Type: new Abstract: We introduce behavioral fidelity -- a third evaluation dimension for synthetic tabular data that measures whether generated data preserves the temporal,

arXiv:2511.17792v2 Announce Type: replace Abstract: While recent video world models can generate highly realistic videos, their ability to perform semantic reasoning and planning remains unclear and u

arXiv:2604.13056v1 Announce Type: new Abstract: This paper presents a practical pipeline for turning text corpora into quantitative semantic signals. Each news item is represented as a full-document e

arXiv:2604.13331v1 Announce Type: new Abstract: In electronic health record (EHR) mining, learning high-quality representations of medical concepts (e.g., standardized diagnosis, medication, and proce

arXiv:2604.13759v1 Announce Type: cross Abstract: Large language model (LLM) agents on multi-step tasks suffer reasoning degradation, looping, drift, stuck states, at rates up to 30% on hard tasks. Cu

arXiv:2604.13051v1 Announce Type: new Abstract: There is debate about whether LLMs can be conscious. We investigate a distinct question: if a model claims to be conscious, how does this affect its dow

Joel Khalili / Wired: The UK unveils Sovereign AI, a £500M fund to invest in domestic AI startups, starting with Callosum, which builds software to help different chips work together — In a bid to min

OpenAI announced the rollout of updates to the Codex desktop application, beginning on the date of the announcement. The updates likely enhance Codex's code generation and AI-assisted programming capa

arXiv:2604.14084v1 Announce Type: new Abstract: On-policy knowledge distillation (OPD) trains a student on its own rollouts under token-level supervision from a teacher. Not all token positions matter

arXiv:2604.13368v1 Announce Type: new Abstract: Fine-tuning large language models (LLMs) aims to adapt pre-trained models to specific tasks using relatively small and domain-specific datasets. Among P

Today we’re announcing Ternary Bonsai: Top intelligence at 1.58 bits Using ternary weights {-1, 0, +1}, we built a family of models that are 9x smaller than their 16-bit counterparts while outperformi

arXiv:2604.13787v1 Announce Type: new Abstract: Large Language Models (LLMs) enhance their problem-solving capability by utilizing external tools. However, in open-world scenarios with massive and evo

arXiv:2603.15620v2 Announce Type: replace Abstract: Vision-Language-Action (VLA) models excel in static manipulation but struggle in dynamic environments with moving targets. This performance gap prim

arXiv:2604.13322v1 Announce Type: new Abstract: Raveling, the loss of aggregates, is a major form of asphalt pavement surface distress, especially on highways. While research has shown that machine le

arXiv:2604.05478v2 Announce Type: replace-cross Abstract: Immune checkpoint inhibitors (ICIs) have transformed cancer therapy; yet substantial proportion of patients exhibit intrinsic or acquired resi

There’s a fault line running through enterprise AI, and it’s not the one getting the most attention. The public conversation still tracks foundation models and benchmarks—GPT versus Gemini, reasoning

arXiv:2604.14116v1 Announce Type: cross Abstract: While Large Language Models (LLMs) have empowered AI research agents to perform isolated scientific tasks, automating complex, real-world workflows, s

arXiv:2505.18232v3 Announce Type: replace-cross Abstract: The deployment of large language models (LLMs) is largely hindered by their large number of parameters. Structural pruning has emerged as a pr

arXiv:2604.13822v1 Announce Type: new Abstract: MLLM-based GUI agents have demonstrated strong capabilities in complex user interface interaction tasks. However, long-horizon scenarios remain challeng

arXiv:2604.13383v1 Announce Type: new Abstract: Ambient Lighting Normalization (ALN) aims to restore images degraded by complex, spatially varying illumination conditions. Existing methods, such as IF

arXiv:2511.23332v2 Announce Type: replace Abstract: Instruction-driven segmentation in remote sensing generates masks from guidance, offering great potential for accessible and generalizable applicati

arXiv:2604.13584v1 Announce Type: new Abstract: We present UNRIO, an uncertainty-aware radar-inertial odometry system that estimates ego-velocity directly from raw mmWave radar IQ signals rather than

arXiv:2604.13928v1 Announce Type: new Abstract: Industrial time-series data from real production environments exhibits substantially higher complexity than commonly used benchmark datasets, primarily

arXiv:2604.06484v2 Announce Type: replace Abstract: Cultural values are expressed not only through language but also through visual scenes and everyday social practices. Yet existing evaluations of cu

arXiv:2604.13596v1 Announce Type: new Abstract: Instance-level object segmentation across disparate egocentric and exocentric views is a fundamental challenge in visual understanding, critical for app

arXiv:2512.14234v2 Announce Type: replace Abstract: Human communication is inherently multimodal and social: words, prosody, and body language jointly carry intent. Yet most prior systems model human

VLM Performance：Qwen3.6 is natively multimodal, and Qwen3.6-35B-A3B showcases perception and multimodal reasoning capabilities that far exceed what its size would suggest, with only around 3 billion a

WAI-ANIMA 1.0 is a newly released Stable Diffusion checkpoint model from the WAI model family, likely combining elements of the WAI-Illustrious anime generation lineage with the Anima diffusion archit

We comprehensively benchmarked Opus 4.7 on document understanding. We evaluated it through ParseBench - our comprehensive OCR benchmark for enterprise documents where we evaluate tables, text, charts,

Anthropic fixed a bug in Claude Opus 4.7 where rate limits for paid subscriptions weren't correctly adjusted for requests using the model's extended context window capabilities. The fix involved reset

We replicated Mythos findings in opencode using public models, not Anthropic's private stack. The moat is moving from model access to validation: finding vulnerability signal is getting cheaper; turni

'We're not looking for the easiest path, we're looking for the right path.' NEW: Aidan Gomez (@aidangomez) is trying to build a very different kind of AI company at Cohere. Based in Toronto, $6.8B-val

Anthropic hosts monthly webinars where they present product updates and new features they've released, with signup information available on their webinars page. The company offers multiple webinar ser

We’ve also added support for 90+ plugins in Codex, giving it more ways to gather context and take action across the tools you already use for docs, project management, code review, creative work, depl

We’ve heard your feedback and we’re working on making it easier to follow everything that’s happening with Claude Code. First, we’re introducing @ClaudeDevs, the official channel to follow for all upd

What you need to know about Opus 4.7 * Takes instructions literally * Better vision means improved computer use and producing slides and other visual artifacts * Optimized for large-scale real-world a

April 2026 brings three major Reinforcement Fine-Tuning updates: Global Training for o4-mini with lower per-token rates across 12+ regions, new GPT-4.1 model graders for richer reward signals, and a c