model releases

TalkPlayData 2: An Agentic Synthetic Data Pipeline for Multimodal Conversational Music Recommendation

arXiv:2509.09685v5 Announce Type: replace-cross Abstract: We present TalkPlayData 2, a synthetic dataset for multimodal conversational music recommendation generated by an agentic data pipeline. In th

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

Testing LLM Arithmetic Reasoning Generalization with Automatic Numeric-Remapping Attacks

DGX agent

arXiv:2606.03606v1 Announce Type: cross Abstract: Large language models achieve strong performance on arithmetic reasoning benchmarks, and one common response to arithmetic brittleness is to delegate

model-releasesarxiv-cs-ai

3 Jun 2026

TeX-1500: A Paired Real-World LWIR Hyperspectral Dataset and Benchmark for Temperature-Emissivity-Texture Decomposition

arXiv:2606.03806v1 Announce Type: new Abstract: Temperature-emissivity-texture (TeX) decomposition seeks to recover object heat state, material spectral response, and visible-like geometric texture fr

The DeepSpeak-Agentic Dataset

arXiv:2606.03686v1 Announce Type: new Abstract: We present DeepSpeak-Agentic, a dataset of videos comprising over 37 hours of semi-structured conversations between a human and an embodied AI agent. We

The Geometry of LLM-as-Judge: Why Inter-LLM Consensus Is Not Human Alignment

model-releasesarxiv-cs-cl

arXiv:2606.03043v1 Announce Type: new Abstract: LMs-as-judges are now standard, yet judges agree strongly with one another while agreeing only weakly with humans. We test whether this reflects shared

The Impact of Configuring Agentic AI Coding Tools on Build-vs-Buy Decisions: A Study Protocol

arXiv:2606.03907v1 Announce Type: cross Abstract: Agentic AI coding tools write code with increasing autonomy and in doing so decide when to import a library and when to implement functionality from s

The Reliability Gap in Benchmark Auditing: Distribution Shift and Scale as Failure Modes of Contamination Detection

arXiv:2606.03305v1 Announce Type: new Abstract: Benchmark contamination, where evaluation examples appear in a model's training data, threatens the validity of LLM assessment. Statistical tools for de

The Ringelmann Effect in Multi-Agent LLM Systems: A Scaling Law for Effective Team Size

arXiv:2606.02646v1 Announce Type: cross Abstract: Inference-time multi-agent LLM scaling lacks a shared unit: counting nominal agents conflates cost with independent evidence. We derive a two-paramete

The Word and the Way: Strategies for Domain-Specific BERT Pre-Training in German Medical NLP

model-releasesarxiv-cs-cl

arXiv:2606.03250v1 Announce Type: new Abstract: Digital healthcare generates vast amounts of clinical text that can support AI-assisted applications, yet German biomedical language models remain limit

This is really intellectually dishonest. I have not been arguing that LLM token prices are increasing (though the all you can eat buffet is …

model-releasesgary-marcus--x

This is really intellectually dishonest. I have not been arguing that LLM token prices are increasing (though the all you can eat buffet is over), I have been arguing the *opposite*, viz that they wil

This story was so implausible that the only way it even (kind of) made sense if it is some sort of internal accounting placeholder at a clou…

model-releasesethan-mollick--x

This story was so implausible that the only way it even (kind of) made sense if it is some sort of internal accounting placeholder at a cloud provider using their own compute. And even then it seems u

thought I was supposed to be the shit-posting account here

model-releasesjerry-liu--x

thought I was supposed to be the shit-posting account here shower thought If: 1. AI is smarter than humans at law, therapy, etc. 2. Humans still like talking to other humans. Then: Humans are just an

ThoughtFold: Folding Reasoning Chains via Introspective Preference Learning

arXiv:2606.03503v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) have achieved remarkable progress thanks to Reinforcement Learning with Verifiable Rewards (RLVR) on Chain-of-Thoughts (Co

Today, we’re excited to introduce Miso One, the most emotive voice model in the world. Miso One is an 8-billion-parameter text-to-speech mod…

model-releasesemad-mostaque--x

Today, we’re excited to introduce Miso One, the most emotive voice model in the world. Miso One is an 8-billion-parameter text-to-speech model for highly expressive speech generation. It emotes like a

Topology-Aware Gaussian Graph Repair for Robust Graph Neural Networks

arXiv:2606.03462v1 Announce Type: new Abstract: Graph neural networks have achieved strong performance on graph-structured data, but their effectiveness depends heavily on the quality of the observed

Towards Characterizing Scientific Image Utility and Upgradability

arXiv:2606.03401v1 Announce Type: new Abstract: Scientific images function as critical evidence in research communication, yet their integrity faces unprecedented threats from AI-generated content tha

Towards Fair Graph Prompting: A Dual-Prompt Mechanism for Mitigating Attribute and Structural Bias

arXiv:2510.23469v2 Announce Type: replace Abstract: Self-supervised pre-training on unlabeled graph data has become a common paradigm for Graph Neural Networks (GNNs). However, an objective gap often

Trading Human Curation for Synthetic Augmentation in RLVR

arXiv:2606.03800v1 Announce Type: cross Abstract: The supply of high-quality training tasks is a central bottleneck for reinforcement learning from verifiable rewards (RLVR) on agentic language models

TriEval: A Resource-Efficient Pipeline for LLM Bias, Toxicity, and Truthfulness Assessment

arXiv:2606.03036v1 Announce Type: new Abstract: LLMs have evolved from basic chatbots to the backbone of the AI ecosystem, now widely used in healthcare, schools, and government services. The domain-w

TSQAgent: Rating Time Series Data Quality via Dedicated Agentic Reasoning

arXiv:2606.03629v1 Announce Type: new Abstract: Assessing the quality of time series (TS) data is fundamental yet inherently challenging due to the multifaceted nature of quality dimensions. Recently,

TurtleAI: Benchmarking Multimodal Models for Visual Programming in Turtle Graphics

arXiv:2606.03626v1 Announce Type: cross Abstract: Vision-language models (VLMs) have been explored for visual programming, where they generate code to solve visual tasks. However, most prior work focu

Uber Caps Usage of AI Tools Like Claude Code to Manage Costs

model-releasessimon-willison

Uber Caps Usage of AI Tools Like Claude Code to Manage Costs I wrote the other day about Uber blowing its 2026 AI budget in four months, and how that wasn't particularly surprising given they would ha

Using open models and inference clouds (which serve open models) is a leading indicator of what is to come. The advantage of open weights is…

model-releasesclem-delangue--x

Using open models and inference clouds (which serve open models) is a leading indicator of what is to come. The advantage of open weights is that you can train, serve, and continually improve your own

v0.30.4: llama-server: fix gemma4 patch wiring (#16477)

model-releasesollama-releases

Ollama v0.30.4 is a patch release addressing a bug in the llama-server component related to incorrect parameter wiring in the Gemma 4 model implementation. This fix ensures Gemma 4 models operate corr

v0.30.4-rc0: Kill llama-server during Windows cleanup (#16458)

model-releasesollama-releases

This release candidate fixes a Windows-specific issue where the llama-server process wasn't being properly terminated during cleanup operations. The fix addresses GitHub issue #16458 and improves the

VidMsg: A Benchmark for Implicit Message Inference in Short Videos

arXiv:2606.03635v1 Announce Type: cross Abstract: Understanding short online videos involves more than identifying visible objects and actions; video makers often include an underlying message or purp

VistaHop: Benchmarking Multi-hop Visual Reasoning for Visual DeepSearch

arXiv:2606.03273v1 Announce Type: cross Abstract: Visual DeepSearch requires multimodal large reasoning model (MLRM) agents to answer complex visual queries by repeatedly inspecting image regions, gro

VLA-Arena: An Open-Source Framework for Benchmarking Vision-Language-Action Models

arXiv:2512.22539v2 Announce Type: replace-cross Abstract: While Vision-Language-Action models (VLAs) are rapidly advancing towards generalist robot policies, it remains difficult to quantitatively und

VLESA: Vision-Language Embodied Safety Agent for Human Activity Monitoring

arXiv:2606.03954v1 Announce Type: new Abstract: As AI systems increasingly assist humans in physical tasks, ensuring safety becomes paramount -- physical actions carry immediate and irreversible conse

vLLM Semantic Router: Signal Driven Decision Routing for Mixture-of-Modality Models

arXiv:2603.04444v3 Announce Type: replace-cross Abstract: As large language models (LLMs) diversify across modalities, capabilities, and cost profiles, the problem of intelligent request routing -- se

WaterSIC: Information-Theoretically (Near) Optimal Linear Layer Quantization

arXiv:2603.04956v2 Announce Type: replace Abstract: This paper considers the problem of converting a given dense linear layer to low precision. The tradeoff between compressed length and output discre

Wavelet Fourier Diffuser: Frequency-Aware Diffusion Model for Reinforcement Learning

arXiv:2509.19305v2 Announce Type: replace-cross Abstract: Diffusion probability models have shown significant promise in offline reinforcement learning by directly modeling trajectory sequences. Howev

We partnered with @FireworksAI_HQ to train open-source models for legal. Here's what we found: 1) Hybrid legal agents can beat frontier mode…

model-releasessonya-huang--x

We partnered with @FireworksAI_HQ to train open-source models for legal. Here's what we found: 1) Hybrid legal agents can beat frontier models on quality and cost by routing selectively to a frontier

We’re bringing new capabilities to GPT-Rosalind, a model series purpose-built for life sciences research at enterprise scale. It brings GPT-…

We’re bringing new capabilities to GPT-Rosalind, a model series purpose-built for life sciences research at enterprise scale. It brings GPT-5.5’s agentic coding and tool use together with stronger int

model-releasesopenai--x

What Benchmarks Don't Measure: The Case for Evaluating Abstention Competence in Autonomous Agents

arXiv:2606.02965v1 Announce Type: new Abstract: Benchmarks for autonomous agents measure whether agents complete tasks, yet this framing is systematically blind to whether an agent should have proceed

What Makes Interaction Trajectories Effective for Training Terminal Agents?

arXiv:2606.03461v1 Announce Type: new Abstract: Stronger code agents are commonly assumed to be superior teachers for post-training, yet this assumption remains poorly disentangled from task difficult

When Model Merging Breaks Routing: Training-Free Calibration for MoE

arXiv:2606.03391v1 Announce Type: cross Abstract: Model merging has emerged as a cost-effective approach for consolidating the capabilities of multiple LLMs without retraining. However, existing mergi

When Should the Teacher Move? Temporal Coupling and Stability in Self On-Policy Distillation

arXiv:2606.03532v1 Announce Type: cross Abstract: Self on-policy distillation trains a student policy against a teacher derived from its own parameter history, yet the teacher's update schedule -- whi

Where Do We (Not) Need Temporal Context in Low-Resource Video Task Adaptation?

arXiv:2606.03837v1 Announce Type: new Abstract: Parameter-efficient fine-tuning (PEFT) and probing enable adaptation of foundation models using only a small number of trainable parameters, making it a

Which Defense Closes Which Threat? Attributing OWASP-LLM-Top-10 Coverage and Its Brittleness Under Paraphrasing

arXiv:2606.02822v1 Announce Type: cross Abstract: Production LLM applications stack several defense families -- refusal-phrase filters, token-budget controls, model allowlists, rate limits, tool-regis

Whose Name Comes Up? II: Benchmarking and Intervention-Based Auditing of LLM-Based Scholar Recommendation

arXiv:2602.08873v2 Announce Type: replace-cross Abstract: Large language models (LLMs) are now used for academic expert recommendation. Existing audits typically evaluate such recommendations in isola

Will Accurate Fields Mislead Photonic Design? FromGlobal Accuracy to Port Readout

arXiv:2606.03038v1 Announce Type: new Abstract: Neural field surrogates can accelerate photonic design loops, but a surrogate that looks accurate in global field error can still mis-rank candidate dev

WISE: A World Knowledge-Informed Semantic Evaluation for Text-to-Image Generation

arXiv:2503.07265v4 Announce Type: replace-cross Abstract: Text-to-Image (T2I) models are capable of generating high-quality artistic creations and visual content. However, existing research and evalua

Wordle 1,809 5/6 🟨⬛🟨⬛⬛ ⬛🟨⬛⬛⬛ ⬛🟩🟨🟨⬛ 🟩🟩🟩🟩⬛ 🟩🟩🟩🟩🟩

model-releasesanthropic--x

This post documents a completed Wordle puzzle (puzzle #1,809) solved in five attempts, showing the progression of letter guesses through color-coded feedback (yellow for correct letters in wrong posit

WRIT: Write-Read Intensive Trajectory Synthesis for Multi-Turn User-Facing Agents

arXiv:2606.02908v1 Announce Type: cross Abstract: Multi-turn user-facing agents must infer user intent from incomplete requests, collect missing information through dialogue and tools, and execute val

3D Segment Anything Model with Visual Mamba for Diagnosing Placenta Accreta Spectrum

arXiv:2606.00489v1 Announce Type: new Abstract: Placenta Accreta Spectrum (PAS) is a rare but highly dangerous obstetric disease. Early and accurate PAS diagnosis is critical for maternal health. Trad

2 Jun 2026

3DCodeBench: Benchmarking Agentic Procedural 3D Modeling Via Code

arXiv:2606.01057v1 Announce Type: cross Abstract: Procedural 3D modeling through code is emerging as a versatile paradigm, offering deterministic, engine-ready, and precisely editable assets that neur

2 Jun 2026

3rd Place at CVPR 2026 CASTLE Challenge: Agentic Multi-View Long-Context Video Understanding via Hierarchical Knowledge Graph Retrieval

arXiv:2606.01933v1 Announce Type: new Abstract: This paper presents our winning methodology for the CASTLE 2026 Challenge at the CVPR 2026 EgoVis Workshop, where our team secured third place globally.