Search: “model-releases”

1. Agreed w @scaling01 that Mythos appears to be better GPT 5.5 on many metrics. 2. Mythos is definitely a major wakeup call wrt security, a…

model-releasesgary-marcus--x

1. Agreed w @scaling01 that Mythos appears to be better GPT 5.5 on many metrics. 2. Mythos is definitely a major wakeup call wrt security, and will pose problems for real-world systems that aren’t wel

Model Releases

A Boundary-Layer Mechanism for One-Third Scaling in Online Softmax Classification

DGX agent

arXiv:2605.22341v1 Announce Type: new Abstract: Hard-label classification is usually trained with smooth surrogate losses, most prominently softmax cross-entropy. We isolate an asymptotic mechanism by

model-releasesarxiv-cs-lg

23 May 2026

A2QTGN: Adaptive Amplitude Quantum-Integrated Temporal Graph Network for Dynamic Link Prediction

arXiv:2605.21916v1 Announce Type: cross Abstract: Dynamic link prediction is important for modeling evolving interactions in complex systems, including social, communication, financial, and transporta

Adaptive RBF-KAN: A Comparative Evaluation of Dynamic Shape Parameters in Kolmogorov-Arnold Networks

arXiv:2605.21534v1 Announce Type: cross Abstract: Kolmogorov-Arnold Networks (KANs) approximate multivariate functions using learnable univariate edge functions, typically parameterized by B-spline ba

Added a DeepSeek Sparse Attention (DSA) from-scratch implementation to my LLMs-from-scratch repo thanks to an awesome new reader contrib. Wi…

model-releasessebastian-raschka--x

Added a DeepSeek Sparse Attention (DSA) from-scratch implementation to my LLMs-from-scratch repo thanks to an awesome new reader contrib. With motivation, overview, and GPT-style model reference imple

Alike Parts: A Feature-Informed Approach to Local and Global Prototype Explanations

arXiv:2605.21646v1 Announce Type: new Abstract: Prototype-based explanations offer an intuitive, example-based approach to support the interpretability of machine learning black box classifiers but of

An entropy formula for the Deep Linear Network

arXiv:2509.09088v3 Announce Type: replace Abstract: We study the Riemannian geometry of the Deep Linear Network (DLN) as a foundation for a thermodynamic description of the learning process. The main

An Improved Adaptive PID Optimizer with Enhanced Convergence and Stability for Deep Learning

arXiv:2605.21968v1 Announce Type: new Abstract: Optimization is essential in deep learning. The foundational method upon which most optimizers are built is momentum-based stochastic gradient descent.

ASSEMBLAGE-DEEPHISTORY: A Cross-Build Binary Dataset with Temporal Coverage

arXiv:2605.21615v1 Announce Type: cross Abstract: Existing binary corpora typically capture only one or two axes of binary variation: they either provide cross-compiler builds without a temporal axis,

AutoBaxBuilder: Bootstrapping Code Security Benchmarking

arXiv:2512.21132v2 Announce Type: replace-cross Abstract: As large language models (LLMs) see wide adoption in software engineering, the reliable assessment of the correctness and security of LLM-gene

Automatic Contextual Audio Denoising

arXiv:2605.22262v1 Announce Type: cross Abstract: Audio context determines which sound components and sources are relevant and which can be perceived as irrelevant (noise) by listeners. For example, t

Billion-Scale Graph Foundation Models

arXiv:2602.04768v2 Announce Type: replace Abstract: Graph-structured data underpins many critical applications. While foundation models have transformed language and vision via large-scale pretraining

Characterizing the Fault Response of the Intel Neural Compute Stick 2 Under Single-Pulse Electromagnetic Fault Injection

arXiv:2605.22437v1 Announce Type: cross Abstract: Vision processing units and other commercial neural-network inference accelerators are increasingly deployed in safety-relevant edge applications, but

Chebyshev Policies and the Mountain Car Problem: Reinforcement Learning for Low-Dimensional Control Tasks

arXiv:2605.22305v1 Announce Type: new Abstract: We analytically solve the Mountain Car problem, a canonical benchmark in RL, and derive an optimal control solution, closing a gap after 36 years. This

ChronoMedicalWorld: A Medical World Model for Learning Patient Trajectories from Longitudinal Care Data

arXiv:2605.21963v1 Announce Type: new Abstract: Long-horizon clinical simulation -- predicting how a patient's physiology evolves over years under specified interventions -- is central to chronic-dise

ChronoVAE-HOPE: Beyond Attention -- A Next-Generation VAE Foundation Model for Specialized Time Series Classification

arXiv:2605.22684v1 Announce Type: new Abstract: Time Series Foundation Models (TSFMs) have become a new component of the state-of-the-art in general time series forecasting. However, adapting them to

Compiling Agentic Workflows into LLM Weights: Near-Frontier Quality at Two Orders of Magnitude Less Cost

arXiv:2605.22502v1 Announce Type: cross Abstract: Agent orchestration frameworks have proliferated, collectively exceeding 290,000 GitHub stars across LangGraph, CrewAI, Google ADK, OpenAI Agents SDK,

Cross-domain benchmarks reveal when coordinated AI agents improve scientific inference from partial evidence

arXiv:2605.22300v1 Announce Type: cross Abstract: Scientific evidence often spans instruments, databases, and disciplines, so no single source records the full phenomenon. This makes it difficult to d

Do Deep Ensembles Actually Capture Uncertainty in Graph Neural Networks?

arXiv:2605.22593v1 Announce Type: new Abstract: While deep ensembles are widely considered to be the default method for uncertainty quantification in deep learning, their effectiveness for graph-struc

Do Not Trust The Auctioneer: Learning to Bid in Feedback-Manipulated Auctions

arXiv:2605.22438v1 Announce Type: cross Abstract: Shilling is the use of artificial bids to make competition appear stronger and push prices upward. We study repeated first-price auctions in which shi

Dropout Universality: Scaling Laws and Optimal Scheduling at the Edge-of-Chaos

arXiv:2605.21648v1 Announce Type: new Abstract: We develop a mean-field theory of dropout as a perturbation of critical signal propagation at the edge of chaos. Dropout shifts the perfect-alignment fi

EmoTrack: Robust Depression Tracking from Counseling Transcripts across Session Regimes

arXiv:2605.22286v1 Announce Type: new Abstract: Text-based counseling is an important interface for AI mental-health support, where transcripts may be used to monitor depression severity and flag sess

Engineering Hybrid Physics-Informed Neural Networks for Next-Generation Electricity Systems: A State-of-the-Art Review

arXiv:2605.21903v1 Announce Type: cross Abstract: The integration of machine learning with domain-specific physics is transforming the design, monitoring, and control of electricity systems, where dat

Evaluation of Pipelines for Data Integration into Knowledge Graphs

arXiv:2605.22304v1 Announce Type: cross Abstract: Integrating new data into knowledge graphs (KG) typically involves different tasks that are executed within workflows or pipelines There are many poss

FD-Bench: A Modular and Fair Benchmark for Data-driven Fluid Simulation

arXiv:2505.20349v2 Announce Type: replace-cross Abstract: Data-driven modeling of fluid dynamics has advanced rapidly with neural PDE solvers, yet a fair and strong benchmark remains fragmented due to

Frequency-Domain Regularized Adversarial Alignment for Transferable Attacks against Closed-Source MLLMs

arXiv:2605.21541v1 Announce Type: cross Abstract: Multimodal large language models (MLLMs) remain vulnerable to transfer-based targeted attacks, where perturbations optimized on open-source surrogate

@GaryMarcus Can't agree more. OpenAI seems struggling to build any successful product after ChatGPT. Codex is promising but it's still early…

model-releasesgary-marcus--x

@GaryMarcus Can't agree more. OpenAI seems struggling to build any successful product after ChatGPT. Codex is promising but it's still early, and it's following Claude Code. It's questionable whether

Google’s new anything-to-anything AI model is wild

model-releasesthe-verge-ai

Last year I deepfaked my kid's stuffed animal to make it look like his plush deer was on vacation. It was an experiment to see if I could re-create the events depicted in a Gemini ad Google was runnin

GPT-5.5 Pro is a very solid fact checker. I can throw entire chapters at it and it will hunt down every key reference accurately. The only r…

model-releasesethan-mollick--x

GPT-5.5 Pro is a very solid fact checker. I can throw entire chapters at it and it will hunt down every key reference accurately. The only real annoyance is that it loves nuance, so returns a lot of “

GraphFlow: A Graph-Based Workflow Management for Efficient LLM-Agent Serving

arXiv:2605.22566v1 Announce Type: new Abstract: Large Language Model (LLM)-based agents demonstrate strong reasoning and execution capabilities on complex tasks when guided by structured instructions,

HIDBench: Benchmarking Large Language Models for Host-Based Intrusion Detection

arXiv:2605.21773v1 Announce Type: cross Abstract: Recent benchmark efforts have advanced the evaluation of large language models (LLMs) in cybersecurity, including tasks such as penetration testing an

Holomorphic Neural ODEs with Kolmogorov-Arnold Networks for Interpretable Discovery of Complex Dynamics

arXiv:2605.22235v1 Announce Type: new Abstract: Complex dynamical systems governed by holomorphic maps such as z^2 + c exhibit fractal boundaries with extreme sensitivity to initial conditions. Accura

How Anthropic's ongoing discussions with the Vatican about ethics and AI led to Christopher Olah being invited to Pope Leo's unveiling of an encyclical on AI (Jack Jenkins/RNS)

Jack Jenkins / RNS: How Anthropic's ongoing discussions with the Vatican about ethics and AI led to Christopher Olah being invited to Pope Leo's unveiling of an encyclical on AI — (RNS) — Pope Leo XIV

model-releasestechmeme

Hugging Face just released @LeRobotHF a humanoid robot you can build for roughly $2,500! A full open stack with hardware, simulation, traini…

model-releasesclem-delangue--x

Hugging Face just released @LeRobotHF a humanoid robot you can build for roughly $2,500! A full open stack with hardware, simulation, training environments, runtime tools, datasets and robot learning

Hybrid Kolmogorov-Arnold Network and XGBoost Framework for Week-Ahead Price Forecasting in Australia's National Electricity Market

arXiv:2605.22387v1 Announce Type: new Abstract: Accurate electricity price forecasting (EPF) is essential for market participants to support operational planning and risk management, yet remains chall

Hyperparameter Transfer with Mixture-of-Expert Layers

arXiv:2601.20205v3 Announce Type: replace Abstract: Mixture-of-Experts (MoE) layers have emerged as an important tool in scaling up modern neural networks by decoupling total trainable parameters from

I called both missteps and that’s part of why OpenAI and their shills hate me. 🤷‍♂️

model-releasesgary-marcus--x

I called both missteps and that’s part of why OpenAI and their shills hate me. 🤷‍♂️ @GaryMarcus It all went wrong with GPT-5 and Sora 2 IMO. Until then OpenAI could do no wrong in the eyes of most peo

I-SAFE: Wasserstein Coherence Metrics for Structural Auditing of Scientific AI Models

arXiv:2605.21731v1 Announce Type: new Abstract: Deep learning models are increasingly used in scientific prediction tasks where strong benchmark performance is often interpreted as evidence of scienti

IKNO: Infinite-order Kernel Neural Operators

arXiv:2605.22182v1 Announce Type: new Abstract: Neural operators have achieved significant success in modern scientific computing due to their flexibility and strong generalization capabilities. Exist

Integrable Elasticity via Neural Demand Potentials

arXiv:2605.22820v1 Announce Type: new Abstract: We propose the Integrable Context-Dependent Demand Network (ICDN), a demand-first neural model for multiproduct retail demand. The model learns log-dema

It was a very good day https://www.spacex.com/launches/starship-flight-12

model-releaseselon-musk--x

SpaceX's Starship Flight 12 achieved significant milestones in its development program, likely including successful booster catch, stage separation, or other critical test objectives. Elon Musk's cele

Learn anything with our new /lesson-generator skill

Learn anything with our new /lesson-generator skill Just released my new /lesson-generator skill. Use it with your agent to learn anything: - generate lessons/courses on any topic - include nano-banan

model-releasesdair-ai--x

Lumberjack: Better Differentially Private Random Forests through Heavy Hitter Detection in Trees

arXiv:2605.22756v1 Announce Type: new Abstract: Random forests are widely used in fields involving sensitive tabular data, but existing approaches to enforcing differential privacy (DP) typically degr

MapTab: Are MLLMs Ready for Multi-Criteria Route Planning in Heterogeneous Graphs?

arXiv:2602.18600v3 Announce Type: replace Abstract: Systematic evaluation of Multimodal Large Language Models (MLLMs) is crucial for advancing Artificial General Intelligence (AGI). However, existing

Measuring Cross-Modal Synergy: A Benchmark for VLM Explainability

arXiv:2605.22168v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) map complex visual inputs to semantic spaces, but interpreting the cross-modal reasoning of VLMs currently relies on pos

Memory-Efficient LLM Pretraining via Minimalist Optimizer Design

arXiv:2506.16659v3 Announce Type: replace Abstract: Training large language models (LLMs) relies on adaptive optimizers such as Adam, which introduce extra operations and require significantly more me

Memory-R2: Fair Credit Assignment for Long-Horizon Memory-Augmented LLM Agents

arXiv:2605.21768v1 Announce Type: new Abstract: Memory-augmented LLM agents enable interactions that extend beyond finite context windows by storing, updating, and reusing information across sessions.

Models Can Model, But Can't Bind: Structured Grounding in Text-to-Optimization

arXiv:2605.21751v1 Announce Type: new Abstract: Text-to-optimization requires two separable capabilities: modeling -- choosing the right optimization structure -- and binding -- grounding every coeffi