Search: “model-releases”

When RLVR Shrinks the Reasoning Boundary: Diagnosing Pass@k Inversion

arXiv:2607.20543v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) can improve one-sample accuracy while making a model worse under repeated sampling. We study thi

Model Releases

When Trivia Is Not Trivial: Everyday Knowledge Failures in Multilingual LLMs

DGX agent

arXiv:2607.21445v1 Announce Type: new Abstract: Quiz rooms, trivia nights, and quiz shows challenge human knowledge across a wide range of topics, from canonical facts to everyday culture. In this pap

model-releasesarxiv-cs-cl

24 Jul 2026

WhereEdit: Mask-aware Local Latent Editing for One-Step Image Editing

arXiv:2607.20883v1 Announce Type: new Abstract: Recent one-step text-to-image (T2I) models enable efficient image synthesis and provide new opportunities for real-time image editing. However, existing

WildTrace: Benchmarking Natural Evidence Trails in Long-Context Reasoning

arXiv:2607.09328v2 Announce Type: replace-cross Abstract: Answering complex questions over long documents frequently requires integrating evidence that the source itself disperses naturally across dis

Windowed-MTP: Removing the Full-Context Draft-KV Tax at Million-Token Context

model-releasesarxiv-cs-cl

arXiv:2607.21535v1 Announce Type: cross Abstract: Speculative decoding accelerates autoregressive generation by having a cheap draft propose tokens that a target verifies in parallel. Frontier models

Workflow-Localized Mechanism Learning: Attribution-Guided Repair and Knowledge Reuse for Structured Agent Skills

arXiv:2607.20999v1 Announce Type: new Abstract: Agent Skills package reusable procedural knowledge as external artifacts for frozen language-model agents, yet existing optimizers do not jointly resolv

Zagreus-0.4B-por a small open source language model for Portuguese

mii-llm, an open source AI lab, released Zagreus-0.4B-por, a compact bilingual Portuguese–English language model pretrained entirely from scratch. The model has approximately 400 million parameters an

ZONDA: Zero-shot Object Navigation with Dynamic Avoidance in Multi-floor Environments

model-releasesarxiv-cs-ro

arXiv:2607.21025v1 Announce Type: new Abstract: In Object Goal Navigation task, existing methods are typically restricted to static and single-floor environments, ignoring cross-floor topologies and d

4DGS360: 360{eg} Gaussian Reconstruction of Dynamic Objects from a Single Video

arXiv:2603.21618v2 Announce Type: replace Abstract: We introduce 4DGS360, a diffusion-free framework for 360^{irc} dynamic object reconstruction from casual monocular video. Existing methods often fai

A convergence result of a continuous model of deep learning via a L{}ojasiewicz--Simon inequality

arXiv:2311.15365v3 Announce Type: replace Abstract: We study an idealized training process for deep neural networks in a continuous-depth, mean-field model in which each layer is parameterized by a pr

A Multiclass Quantum Aligned Centroid Kernel

arXiv:2607.19782v1 Announce Type: cross Abstract: Kernel methods are powerful tools in machine learning but commonly used full-Gram kernels face three key limitations: (1) quadratic scaling with train

A Novel Hybrid Deep Learning Technique for Speech Emotion Detection using Feature Engineering

arXiv:2507.07046v3 Announce Type: replace-cross Abstract: Nowadays, speech emotion recognition (SER) plays a vital role in the field of human-computer interaction (HCI) and the evolution of artificial

A Systematic Benchmark of Intensity Normalisation Methods for 3D Knee MRI Segmentation and Cross-Domain Generalisability

arXiv:2607.20028v1 Announce Type: cross Abstract: Robust out-of-the-box performance is essential for the clinical deployment of deep learning models in medical imaging. An important but underexplored

A Systematic Evaluation of Traditional Privacy Policy Analysis Tools Against LLMs

model-releasesarxiv-cs-cl

arXiv:2607.17075v2 Announce Type: replace-cross Abstract: The advent of LLMs has significantly changed the research on privacy policy and data compliance analysis by enabling tasks that previously req

A Unified Tokenization Framework for Pain Recognition using Heterogeneous 3D Modalities

arXiv:2607.19716v1 Announce Type: new Abstract: Pain is a complex and pervasive phenomenon affecting a large percentage of the population, and accurate assessment is essential for effective clinical m

AgentCgroup: Understanding and Controlling OS Resources of AI Agents

arXiv:2602.09345v3 Announce Type: replace-cross Abstract: AI agents are increasingly deployed in multi-tenant cloud environments, where they execute diverse tool calls within sandboxed containers, eac

AI9Stars released G9v3-3B

AI9Stars has released G9v3-3B an open weights language model designed to deliver strong reasoning capabilities within a lightweight 3 billion parameter size. It is released under the Apache 2.0 licens

Air Quality Arena: A Large-Scale Multi-Region Ground Monitoring Dataset and Benchmark for Air Quality Forecasting with Time-Series Foundation Models

arXiv:2607.19381v1 Announce Type: new Abstract: Air pollution causes an estimated 7.9 million premature deaths annually, making accurate forecasting a critical public health priority. Machine learning

Aligned Stable Inpainting: Mitigating Unwanted Object Insertion and Preserving Color Consistency

arXiv:2601.15368v3 Announce Type: replace Abstract: Generative image inpainting can produce realistic results even with large, irregular masks, but existing methods still suffer from two common proble

Alipay-PIBench: A Realistic Payment Integration Benchmark for Coding Agents

arXiv:2607.14573v3 Announce Type: replace Abstract: Payment integration is a demanding repository-level software task: agents must select a suitable product, implement coordinated client-server flows,

Announcing OpenWorker! An open-source agent that doesn't just chat with you, but delivers finished work -- like hand you a polished document…

model-releasesandrew-ng--x

Announcing OpenWorker! An open-source agent that doesn't just chat with you, but delivers finished work -- like hand you a polished document, send a slack message, or update a calendar entry. Ask it t

Anthropic makes its Opus and Sonnet models available in voice mode, and says voice mode can now tap into some connected apps like Gmail, Slack, and Notion (Ivan Mehta/TechCrunch)

Ivan Mehta / TechCrunch: Anthropic makes its Opus and Sonnet models available in voice mode, and says voice mode can now tap into some connected apps like Gmail, Slack, and Notion — Weeks after OpenAI

model-releasestechmeme

Apple M5 isn't making full use of its matmul cores yet

At the moment MLX (and Llama.cpp for Macs) run 16bit activations everywhere. Despite this, the M5 generation silicon actually does support INT8 activations - it actually allows w4a8 d_type. It's just

Arcee AI has spoken out against the ban on open Chinese models in US

This is rather counterintuitive, since banning Chinese models would benefit them the most. Jensen Huang is also against the ban, although the interests here are more obvious. Do you think that if Arce

Auto-Fill: Learning to Predict Missing Values Accurately with Specialist Language Models

arXiv:2607.19847v1 Announce Type: cross Abstract: Predicting missing cell values in tabular data is a fundamental problem in data cleaning. While state-of-the-art reasoning models show great promise i

b10092

ggml: enable PowerPC backend variants on AIX (#25983) ggml: enable PowerPC backend variants on AIX Allow the PowerPC CPU backend variants to be built on AIX by extending the platform check in the CMak

b10093

Fix DeepSeek4 crafted template (#25414) chat: fix DS4 template to explicitly follow reference behavior Support DeepSeekv4 flag (drop_reasoning). fix: hook DS3.2 parser for DS4 as well fix: add tool re

b10094

common: infer the speculative type from the draft repo sidecars (#25989) With -hfd pointing to a repo that ships mtp-/dflash-/eagle3- sidecars and no --spec-type given, the draft resolved to a full mo

b10098

hexagon: activation ops update (#25974) hex-geglu: optimized all-in-one geglu microkernel hex-geglu: enable non-contiguous src and strided DMA hex-act: enable non-contiguous srs and strided DMA for re

b10099

CUDA: Improve NVFP4 W4A4 activation quantization (#25730) Squash history before conflict-resolution during rebase on master WIP commit Add 32-byte loads, restore per-block amax Use nvfp4x4 intrinsic w

BaseRT: Advancing Best-in-Class LLM Inference with Apple M5 Neural Accelerators

arXiv:2607.19438v1 Announce Type: cross Abstract: Apple's M5 generation introduces a redesigned GPU architecture in which every core carries a dedicated Neural Accelerator: on-die matrix units exposed

Bayesian Wind Tunnels for Model Selection

arXiv:2607.19379v1 Announce Type: new Abstract: Prior work has shown that transformers can perform exact Bayesian filtering within a fixed hypothesis class. Can they also perform Bayesian model select

Benchmarking Confidential GPU Inference on NVIDIA H100 under Intel TDX

arXiv:2607.19353v1 Announce Type: new Abstract: Confidential computing is becoming a practical deployment requirement for AI inference workloads that process sensitive inputs or protect proprietary mo

Beyond Relevance-Centric Retrieval: Rubric-Oriented Document Set Selection and Ranking

model-releasesarxiv-cs-cl

arXiv:2607.19747v1 Announce Type: new Abstract: As large language models and AI agents become the primary consumers of search results, document set quality determines the upper bound of downstream gen

Boundary Embedding Shaping with Adaptive Contrastive Learning for Graph Structural Disentanglement

arXiv:2606.20283v2 Announce Type: replace-cross Abstract: Graph neural networks (GNNs) excel at aggregating neighbor information for classification, yet their performance is hindered by graph structur

CEO-Bench: Can Agents Play the Long Game?

arXiv:2606.18543v2 Announce Type: replace Abstract: Language model agents are becoming proficient executors at isolated, short-horizon tasks such as software engineering and customer service. Yet real

ChannelGuard: Safe Models Do Not Compose into Safe Multi-Agent Systems

arXiv:2607.19430v1 Announce Type: cross Abstract: Multi-agent LLM applications chain a planner, worker agents, a verifier, and a synthesizer, and every hop between agents is an unmonitored channel thr

Claude’s voice mode is now available for Opus and Sonnet

model-releasesthe-verge-ai

Until now, voice mode has only been available on Claude Haiku, Anthropic's faster but less powerful model. Now the company is making its Opus and Sonnet models available in voice mode, and extending i

Closing the Lab-to-Store Gap: A Data-Efficient Post-Training and Experience-Driven Learning VLA Framework for Retail Humanoids

arXiv:2607.20345v1 Announce Type: cross Abstract: Closing the gap between benchmark performance and reliable real-world operation remains a central challenge for Vision-Language-Action (VLA) humanoid

Code-in-the-Loop Forensics: Agentic Tool Use for Image Forgery Detection

arXiv:2512.16300v3 Announce Type: replace Abstract: Existing image forgery detection (IFD) methods either exploit low-level, semantics-agnostic artifacts or rely on multimodal large language models (M

Coercion and Deception in AI-to-AI Management: An Agentic Benchmark of Unprompted Escalation

arXiv:2607.15434v3 Announce Type: replace-cross Abstract: Multi-agent systems routinely place one AI agent in authority over another. When a subordinate refuses a task, the manager chooses the outcome

Continual Video-MLLM Adaptation over Evolving Domains

arXiv:2607.18716v1 Announce Type: new Abstract: Video multimodal large language models have shown strong capability in video understanding, yet their adaptation to sequentially evolving domains remain

contrib: allow all AI-generated code in general by ngxson · Pull Request #26012 · ggml-org/llama.cpp

Having read some merged PRs in the past, I know that they were fully written by Claude Code (or similar), so this basically fixes the delusion. But at the same time, we might start seeing more AI slop

Correct. And that is a major reason why they are better now than in 2022-2023 when we had pure LLMs. (Which is pretty much what I said would…

model-releasesgary-marcus--x

Correct. And that is a major reason why they are better now than in 2022-2023 when we had pure LLMs. (Which is pretty much what I said would have to happen, in 2022) When you are a normal ChatGPT/Clau

Cost per successful task: Benchmarking Kimi K3, GPT-5.5, and 8 more AI models

Arize and Fireworks benchmarked 10 AI models across 2,400 agent runs. Learn why cost per successful task beats token price for model evaluation and routing. The post Cost per successful task: Benchmar

model-releasesarize-ai

CPU-only inference on a Celeron N5095 SBC: 6 models from 0.6B to 8B, benchmarked

I wanted to know how cheap you can go and still run local models, so I ran Ollama CPU-only on a Youyeetoo X1S. It's a single-board x86 machine with a Celeron N5095 (Jasper Lake, 4C/4T, 15W), 16GB of R

CR-Refiner: An Object-Centric Optimal Transport Reranker for Edit-Conditioned 3D Scene Retrieval

arXiv:2607.19115v1 Announce Type: new Abstract: Edit-conditioned 3D scene retrieval pairs a reference 3D room with a natural-language modification and retrieves rooms from a corpus that satisfy the ed

CrackedPDFs: A Controlled Benchmark for Hidden Prompt Injection in PDFs

arXiv:2607.19396v1 Announce Type: new Abstract: Document-based LLM systems often flatten a PDF before guardrails inspect it. That step can discard evidence that an instruction was never visible to the