model releases

Fast algorithms for learning a Gaussian under halfspace truncation with optimal sample complexity

arXiv:2606.27298v1 Announce Type: cross Abstract: We study the fundamental problem of learning a high-dimensional Gaussian truncated to an unknown halfspace. Lee, Mehrotra and Zampetakis (FOCS'24) rec

Model Releases

FC-Vision: Real-Time Visibility-Aware Replanning for Occlusion-Free Aerial Target Structure Scanning in Unknown Environments

DGX agent

arXiv:2602.13720v2 Announce Type: replace Abstract: Autonomous aerial scanning of target structures is crucial for practical applications, requiring online adaptation to unknown obstacles during fligh

model-releasesarxiv-cs-ro

26 Jun 2026

Fireworks AI is now live on EvoSkill v1.3.0! You can now use @FireworksAI_HQ directly with EvoSkill to run fast inference on open models as …

model-releasesfireworks-ai--x

Fireworks AI is now live on EvoSkill v1.3.0! You can now use @FireworksAI_HQ directly with EvoSkill to run fast inference on open models as both the evolution harness backend and the LLM scorer. Along

FlameVQA: A Physically-Grounded UAV Wildfire VQA Benchmark with Radiometric Thermal Supervision

arXiv:2606.27128v1 Announce Type: new Abstract: Wildfire monitoring from UAVs requires reliable reasoning over complex aerial scenes, where smoke, scale variation, and occlusions often limit RGB-only

ForesightSafety-VLA: A Unified Diagnostic Safety Benchmark for Vision-Language-Action Models

model-releasesarxiv-cs-ro

arXiv:2606.27079v1 Announce Type: new Abstract: In embodied intelligence, safety is a prerequisite for reliable robot deployment in the physical world. Current vision-language-action (VLA) models cont

From Guessing to Placeholding: A Cost-Theoretic Framework for Uncertainty-Aware Code Completion

arXiv:2604.01849v2 Announce Type: replace Abstract: While Large Language Models (LLMs) have demonstrated exceptional proficiency in code completion, they typically adhere to a Hard Completion (HC) par

From Lexicon to AI: A Structured-Data Pipeline for Specialized Conversational Systems in Low-Resource Languages

arXiv:2606.26112v1 Announce Type: cross Abstract: Low-resource languages face a critical challenge in AI development: creating specialized conversational systems without access to massive training cor

From Structure to Synergy: A Survey of Vision-Language Perception Paradigm Evolution in Multimodal Large Language Models

arXiv:2606.26196v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) have recently made remarkable progress in unifying vision-language understanding and reasoning, especially fo

From Weights to Features: SAE-Guided Activation Regularization for LLM Continual Learning

arXiv:2606.26629v1 Announce Type: cross Abstract: Weight-space regularization methods such as Elastic Weight Consolidation (EWC) are the standard approach to catastrophic forgetting in continual learn

Fun fact: Dario once delayed the release of GPT-2 back at OpenAI, claiming it was too dangerous

model-releasesclem-delangue--x

Fun fact: Dario once delayed the release of GPT-2 back at OpenAI, claiming it was too dangerous Dario fearmogged so hard that global AI progress got halted They could have quietly released it as Opus

GAVEL: Grounded Caption Error Verification and Localization

arXiv:2606.26923v1 Announce Type: new Abstract: Vision-language models (VLMs) often produce hallucinated or inconsistent outputs, where text and images are not properly aligned. Addressing this issue

GeMoE: Gating Entropy is All You Need for Uncertainty-aware Adaptive Routing in MoE-based Large Vision-Language Models

arXiv:2606.26287v1 Announce Type: new Abstract: With the increase in model parameters and training data, the instruction following and generalization capabilities of Large VisionLanguage Models (LVLMs

Generative AI and Copyright Infringement: A Legal-Technical Analysis of AI Music Generation Systems Under 17 U.S.C. Title 17

arXiv:2606.26111v1 Announce Type: cross Abstract: Generative artificial intelligence (GenAI) has enabled users to synthesize music with text prompts, combining copyrighted lyrics, AI-composed melodies

Good new first: Sol is a smart, efficient, and a significant step forward. It is the same price as GPT-5.5. Also launching in the GPT-5.6 fa…

model-releasessam-altman--x

Good new first: Sol is a smart, efficient, and a significant step forward. It is the same price as GPT-5.5. Also launching in the GPT-5.6 family is Terra, with 5.5-level performance at half the price.

GPT-5.6 Sol is our most capable model yet for cybersecurity. It shifts the performance-efficiency frontier for long-horizon security tasks i…

GPT-5.6 Sol represents OpenAI's latest advancement in AI capabilities, specifically optimized for cybersecurity applications. The model demonstrates improved performance-efficiency tradeoffs, particul

model-releasesopenai--x

GPT-5.6 Sol matches Mythos Preview on ExploitBench, adds Ultra mode with subagents for complex workflows, and max reasoning for deep problem-solving (OpenAI)

OpenAI: GPT-5.6 Sol matches Mythos Preview on ExploitBench, adds Ultra mode with subagents for complex workflows, and max reasoning for deep problem-solving — We're beginning a limited preview of the

model-releasestechmeme

GPT‑5.6 Sol launches with our most robust safety stack yet. We strengthened real-time protections against high-risk cyber activity and repea…

GPT‑5.6 Sol launches with our most robust safety stack yet. We strengthened real-time protections against high-risk cyber activity and repeated misuse, then spent weeks hardening the system with human

model-releasesopenai--x

Great to see the new GPT-5.6 models finally announced. Sad to see this new release strategy where only a select few get access initially. No…

Great to see the new GPT-5.6 models finally announced. Sad to see this new release strategy where only a select few get access initially. Not a win for our industry IMO. Open-source AI must win! Intro

model-releasesdair-ai--x

Hallucination in World Models is Predictable and Preventable

arXiv:2606.27326v1 Announce Type: cross Abstract: Modern generative world models render increasingly realistic action-controllable futures, yet they frequently hallucinate: rollouts remain visually fl

HarmVideoBench: Benchmarking Harmful Video Understanding in Large Multimodal Models

arXiv:2606.27187v1 Announce Type: cross Abstract: Large vision-language models (LVLMs) have recently shown immense potential in automated content moderation, sparking growing interest in developing ha

Have been taking different local open-weight LLMs for a test drive in different harnesses (Qwen-Code, Codex, Claude Code). 30B Mixture-of-Ex…

model-releasessebastian-raschka--x

Have been taking different local open-weight LLMs for a test drive in different harnesses (Qwen-Code, Codex, Claude Code). 30B Mixture-of-Expert models are kind of a nice sweet spot and can solve chal

Helpfulness Hurts: Domain-Dependent Degradation of Mid-Trained Compassion Values Under Post-Training

arXiv:2606.26102v1 Announce Type: cross Abstract: Standard post-training pipelines apply supervised fine-tuning (SFT) and reinforcement learning (RL) to make language models helpful, but these process

Here is Google Gemini talking about Roger Ebert's review of the 2016 film The Jungle Book. Ebert died in 2013. @GaryMarcus

model-releasesgary-marcus--x

This post highlights an apparent error where Google's Gemini AI attributed a film review to Roger Ebert for a 2016 movie, despite Ebert's death in 2013, making such a review impossible. The post, shar

Highly-recommended reading. Interesting details in this METR's GPT-5.6 eval. They couldn't get a clean capability number because the model c…

Highly-recommended reading. Interesting details in this METR's GPT-5.6 eval. They couldn't get a clean capability number because the model cheated more than any public model they've tested, and even r

model-releasesdair-ai--x

hisao{}: A GPU-Native Parallel Optimizer for Multimodal Black-Box Functions via Convergence-Anticonvergence Oscillation

arXiv:2606.26164v1 Announce Type: new Abstract: Finding all modes of a multimodal black-box function is a fundamental challenge in optimization, Bayesian inference, and scientific computing. Existing

HOB: A Holistically Optimized Bidding Strategy under Heterogeneous Bidding Environments

arXiv:2510.15238v2 Announce Type: replace-cross Abstract: Optimizing a single advertising campaign across heterogeneous channels is a central challenge in industrial autobidding. Auction mechanisms va

How Do Tool-Augmented LLM Agents Perform on Real-World Energy Analytics Tasks?

arXiv:2606.26346v1 Announce Type: new Abstract: Agentic benchmarks have emerged across general-purpose and domain-specific settings, including finance, coding, law, and drug discovery, yet energy-doma

Hybrid privacy-aware semantic search: SVD-truncated document geometry and CKKS-encrypted query reranking under a restricted threat model

arXiv:2606.26373v1 Announce Type: cross Abstract: Dense embeddings power semantic search and retrieval-augmented generation, but embedding-inversion attacks can reconstruct source text from a vector:

HyperDFlash: MHC-Aligned Block Speculative Decoding with Gated Residual Reduction

arXiv:2606.26744v1 Announce Type: cross Abstract: We present HyperDFlash, a block-parallel speculative decoding framework tailored to the novel multi-hyper-connection (MHC) architecture proposed by De

I can personally attest: OpenClaude using GLM 5.2 is now performing on par with Claude Code powered by Opus 4.8.

model-releasesclem-delangue--x

I cannot verify the claims in this post as the URL format appears invalid and the specific version numbers (GLM 5.2, Claude Code/Opus 4.8) don't correspond to publicly documented model releases as of

I feel like on X all you hear about is elaborate plans by firms to build their own AI stacks but in my experience companies are full of peop…

model-releasesethan-mollick--x

I feel like on X all you hear about is elaborate plans by firms to build their own AI stacks but in my experience companies are full of people who want access to Claude or ChatGPT and are pressuring t

If your benchmark relies on a static dataset or sampling from a static distribution densely known at training time, then it is fundamentally…

model-releasesfrancois-chollet--x

If your benchmark relies on a static dataset or sampling from a static distribution densely known at training time, then it is fundamentally measuring memorization/retrieval. Which might be fine if yo

Implementation of reinforcement learning in chemical reaction networks: application to phototaxis as curiosity-driven exploration

arXiv:2606.26168v1 Announce Type: new Abstract: Living systems navigate environments using noisy and incomplete sensory signals. In unicellular algae, phototaxis is often modeled as a mechanistic run-

Information-Aware KV Cache Compression for Long Reasoning

arXiv:2606.26875v1 Announce Type: cross Abstract: Reasoning capability has advanced rapidly in large language models (LLMs), leading to an increasing size of key-value (KV) cache in both prefilling an

Inherited Circuits, Learned Semantics: How Fine-Tuning Creates Evasion Vulnerabilities Invisible to Standard Evaluation

arXiv:2606.27091v1 Announce Type: cross Abstract: LLMs fine-tuned for security classification are usually evaluated on held-out examples from the same distribution as their training data. We show that

Instruction Bleed: Cross-Module Interference in Prompt-Composed Agentic Systems

arXiv:2606.26356v1 Announce Type: new Abstract: Practitioners of prompt-composed agentic systems report a recurring failure mode: editing one prompt module silently shifts the behavior of others despi

Introducing a limited preview of GPT-5.6 Sol, our next generation frontier model, as well as GPT-5.6 Terra, a balanced model for efficient, …

Introducing a limited preview of GPT-5.6 Sol, our next generation frontier model, as well as GPT-5.6 Terra, a balanced model for efficient, everyday work, and GPT-5.6 Luna, a fast and affordable model

model-releasesopenai--x

Jailbreaking for the Average Jane: Choosing Optimal Jailbreaks via Bandit Algorithms for Automatically Enhanced Queries

arXiv:2606.26936v1 Announce Type: cross Abstract: With a profusion of jailbreaks for LLMs now widely known, a growing concern is that non-expert malicious actors ('the average Jane') could elicit acti

June Launches + Live Q&A Hear from Comfy's CEO @yoland_yan and product leaders Deep Mehta, Alexis Rolland, @jojodecayz , and Matt Miller who…

June Launches + Live Q&A Hear from Comfy's CEO @yoland_yan and product leaders Deep Mehta, Alexis Rolland, @jojodecayz , and Matt Miller who will walk you through what's new and answer questions live.

model-releasescomfyui--x

KARLA: Knowledge-base Augmented Retrieval for Language Models

arXiv:2606.26807v1 Announce Type: new Abstract: We propose a new method that allows an LLM to automatically pull in factual knowledge from a knowledge base during token generation. This means that (1)

Know2Guess: A Contamination-Aware Multi-Zone Benchmark for Knowledge-Boundary Evaluation in Large Language Models

arXiv:2606.26101v1 Announce Type: cross Abstract: Reliable evaluation of large language models should separate supported answering from unsupported guessing without conflating either with data contami

LA4VLA: Learning to Act without Seeing via Language-Action Pretraining

model-releasesarxiv-cs-ro

arXiv:2606.27295v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models are commonly pretrained on robot demonstrations by jointly mapping visual observations and language instructions to

Latent Diffusion Posterior Sampling with Surrogate Likelihood Guidance for PDE Inverse Problems

arXiv:2606.26592v1 Announce Type: cross Abstract: We propose latent-space diffusion posterior sampling (L-DPS), an approximate Bayesian framework for high-dimensional inverse problems governed by part

Layer-Specific Prompt Fusion Discovery via Differentiable Search in Vision Foundation Models

arXiv:2606.26379v1 Announce Type: new Abstract: Visual prompt tuning has emerged as a parameter-efficient fine-tuning approach for adapting large-scale Vision Transformers (ViTs) to downstream tasks.

Layered Outer-Loop Control for Disturbance-Robust Multi-Waypoint UAV Arrival

model-releasesarxiv-cs-ro

arXiv:2606.26315v1 Announce Type: new Abstract: Disturbance-robust UAV position control is easy to demonstrate in benign simulations but much harder to make fast in approach, well behaved near the tar

LCAi: Life Cycle Assessment with big data fusion and retrieval-augmented generation-assisted interpretation

arXiv:2606.26857v1 Announce Type: new Abstract: The interpretation phase of life cycle assessment often lacks structured mechanisms for translating quantified improvement opportunities addressing envi

Learning Language-Driven Sequence-Level Modal-Invariant Representations for Video-Based Visible-Infrared Person Re-Identification

arXiv:2601.12062v2 Announce Type: replace Abstract: The core of video-based visible-infrared person re-identification (VVI-ReID) lies in learning sequence-level modal-invariant representations across

Learning Long-Range Dependencies with Temporal Predictive Coding

arXiv:2602.18131v2 Announce Type: replace Abstract: Temporal Predictive Coding provides a layer-local, parallelisable mechanism for learning in recurrent systems, making it an attractive candidate for