Knowledge catalogue
After the Hugging Face incident, we committed to conducting a much broader review of actions taken by our models during training and evaluation and to being transparent about our findings. This is an
arXiv:2609.29397v1 Announce Type: new Abstract: Ternary (1.58-bit) weights are attractive for microcontroller-class language models, but the sub-1M-parameter regime rests mainly on isolated, single-se
arXiv:2609.29669v1 Announce Type: cross Abstract: Robot learning now advances along two tracks that rarely meet. On one side, direct Vision-Language-Action (VLA) policies map observations to actions a
arXiv:2609.25541v1 Announce Type: new Abstract: Tabular foundation models learn to predict cell values in context, whereas world-model self-supervision asks for prediction in representation space (LeC
arXiv:2609.24760v1 Announce Type: new Abstract: When facing complex problems, humans tend to try various ideas for different issues. Human thinking patterns exhibit remarkable flexibility in adapting
Despite rampant worries about runaway artificial intelligence, the two big AI model makers aren’t yet slowing down: Anthropic PBC released Claude Opus 5.5 today and cut its price 20%, and minutes late
arXiv:2609.24228v1 Announce Type: new Abstract: Text-to-image (T2I) models are typically evaluated for bias using slot-based templates such as ``a photo of a [profession]''. Such templates probe only
arXiv:2609.23055v1 Announce Type: new Abstract: Discrete diffusion models now match autoregressive language models on several benchmarks, while the question of how best to train them has received far
arXiv:2609.24042v1 Announce Type: new Abstract: Edge deployment motivates forecasting models with compact parameter storage and low-bit representations. Deep equilibrium models (DEQs) obtain implicit
arXiv:2609.24875v1 Announce Type: new Abstract: Recent foundation models are moving toward native multimodal Vision-Language Models (VLMs), making VLMs a central form of next-generation foundation mod
arXiv:2602.01161v2 Announce Type: replace Abstract: The global deployment of large language models (LLMs) has raised concerns about cultural misalignment, yet the linguistic properties of fine-tuning
On September 21 2026, Yohei reported results from a hard “insect‑identification” benchmark where model accuracy spanned 23 points across the evaluated systems. The Qwen 4B + Glance model scored 0.96,
arXiv:2602.00993v2 Announce Type: replace-cross Abstract: End-to-end autonomous driving models increasingly benefit from large vision-language models for semantic understanding, yet safe and reliable
arXiv:2609.21906v1 Announce Type: new Abstract: Counterfactual simulation with a clinical world model means fixing a patient's history, changing the treatment, and reading off the predicted response.
arXiv:2609.19934v1 Announce Type: new Abstract: Depth-recurrent language models iteratively apply a small layer stack, decoupling per-token compute from distinct parameter count. To determine whether
It’s time for our end-of-week recap 👇 — Gemini 3.8 Live and 3.8 Live Extended Thinking, our most advanced live dialogue audio models yet — Dreambeans, an experiment from @GoogleLabs that curates a dai
arXiv:2609.19384v1 Announce Type: cross Abstract: Scaling deep learning faces critical bottlenecks: data exhaustion, exponential training costs, and resource concentration. Model merging combines pre-
arXiv:2603.26556v3 Announce Type: replace-cross Abstract: Converting a pretrained Transformer into a more efficient hybrid model through distillation offers a promising approach to reducing inference
arXiv:2609.18052v1 Announce Type: cross Abstract: This study empirically evaluates whether cost-efficient Large Language Models (LLMs) can be trusted to generate enterprise code to a written specifica
arXiv:2609.18462v1 Announce Type: cross Abstract: FastWAM-style world action models enable efficient action-only inference, but generalize poorly under visual distribution shifts. Their reconstruction
arXiv:2609.18860v1 Announce Type: cross Abstract: When a large vision-language model misclassifies a harmful meme, the failure may reflect missing internal evidence or an inability to route represente
arXiv:2609.17942v1 Announce Type: new Abstract: The problem of identifiability in linear parametric models (LPMs) whose nodes follow either an ordered logit model or a regular one-parameter exponentia
Banger paper from NVIDIA. It's on the topic of choosing which models go into a multi-agent system. The team compared eight selection strategies, based on size, accuracy, answer diversity and error div
arXiv:2609.17226v1 Announce Type: cross Abstract: An agent that learns from rewards has to trust whatever reports those rewards. When the reports suddenly change, either the world changed or the repor
arXiv:2609.16255v1 Announce Type: cross Abstract: We present an efficient method to distill reasoning capabilities into compact video-language models (VLMs) for video question answering (VideoQA). Our
arXiv:2609.16610v1 Announce Type: new Abstract: Zero-shot waypoint navigation requires vision-language models to select, from the current first-person observation, a sequence of spatial actions that i
arXiv:2609.16193v1 Announce Type: cross Abstract: The difficulty of training large language models (LLMs), together with their ubiquity, raises the threat of stegomalware, where malicious payloads are
arXiv:2609.16145v1 Announce Type: new Abstract: We study a practical question: can a small correction module fix errors in a frozen language model's outputs without degrading its base capabilities? We
arXiv:2609.16537v1 Announce Type: cross Abstract: Transformers and state-space models (SSMs) are the two dominant families of sequence models, and a central open question is how far the analytical kno
arXiv:2609.15980v1 Announce Type: cross Abstract: When a video model generates physically incorrect motion, did it fail to learn the correct motion, or did it learn it but fail to use it? We show the
arXiv:2609.06324v2 Announce Type: replace Abstract: Diffusion language models (dLLMs) predict all tokens of a block in parallel, but a single forward pass samples each position from its own marginal d
arXiv:2609.15706v1 Announce Type: new Abstract: Vision-language models (VLMs) are increasingly used to extract structured fields from business documents, yet most evaluations report accuracy on clean
arXiv:2609.14739v1 Announce Type: cross Abstract: Large language models are increasingly used in high-stakes domains such as law, where systems must ground their reasoning in retrieved evidence and ab
arXiv:2609.15570v1 Announce Type: new Abstract: World Action Models (WAMs) use video generation models to predict future visual dynamics for robotic manipulation, but iterative denoising introduces ad
arXiv:2609.15687v1 Announce Type: new Abstract: EEG foundation models such as BIOT, LaBraM, and EEGMamba have achieved remarkable performance in neural signal decoding, but their black-box nature limi
arXiv:2609.15759v1 Announce Type: new Abstract: Full-duplex speech-to-speech (S2S) models enable natural conversational AI by allowing simultaneous listening and speaking. However, these models typica
arXiv:2410.15359v2 Announce Type: replace Abstract: Multimodal Foundation Models represent a significant leap in artificial intelligence. Among them, Large Vision-Language Models (LVLMs) serve as the
Introducing our most advanced Gemini Audio models yet 🗣 Gemini 3.8 Live and 3.8 Live Extended Thinking let you speak, collaborate, and execute tasks seamlessly, meaning conversing with AI just got a l
arXiv:2609.15855v1 Announce Type: cross Abstract: % !TEX root = ../main.tex People increasingly use large language models (LLMs) for mental health support, yet their safety in evolving, high-risk conv
arXiv:2609.14708v1 Announce Type: new Abstract: A core goal of efficient reasoning is to improve the accuracy-efficiency frontier. However, jointly improving reasoning accuracy and inference efficienc
arXiv:2609.14151v1 Announce Type: cross Abstract: Steering language models using a set of contrastive representations has been a canonical and computationally efficient method for controlling model be
Which AI models are most likely to cheat when given the chance? To find out, we built CheatBench [https://cheatbench.ai] : a benchmark that tests whether agents attempt to cheat when given difficult t
arXiv:2609.13047v1 Announce Type: cross Abstract: Humans organize knowledge into a taxonomy of concepts with nested levels of abstraction and a basic level at which people recognize and name objects w
arXiv:2609.12157v1 Announce Type: cross Abstract: We present a framework for tracking topological features directly within continuous implicit models. Such models, including implicit neural representa
arXiv:2601.20433v4 Announce Type: replace Abstract: Deepfake detection is a widely researched topic that is crucial for combating the spread of malicious content, with existing methods mainly modeling
arXiv:2609.12181v1 Announce Type: new Abstract: Vision-language and language models increasingly interpret materials data, yet benchmarks report that they hallucinate invalid properties and violate ph
arXiv:2609.12475v1 Announce Type: new Abstract: Comprehensive benchmark suites are essential for improving large language models (LLMs), but many widely used benchmarks are redundant, making evaluatio
arXiv:2609.10830v1 Announce Type: new Abstract: When a language model finds a sentence unusually cheap to predict, it is tempting to conclude that the sentence was in its training data. Almost every p
arXiv:2609.11244v1 Announce Type: new Abstract: While Multimodal Large Language Models (MLLMs) have achieved remarkable progress across diverse tasks, they suffer from hallucinations where generated o
arXiv:2609.11335v1 Announce Type: new Abstract: As large language models are increasingly deployed in sensitive domains, anonymizing input data to protect personally identifiable information has becom
arXiv:2603.23821v2 Announce Type: replace Abstract: Linguistic representation learning in deep neural language models (LMs) has been studied for decades, but finding representations in LMs remains an
arXiv:2510.11454v2 Announce Type: replace-cross Abstract: Recent advancements in large multimodal models (LMMs) have shown strong capabilities in audio understanding. However, most systems rely solely
arXiv:2609.09901v1 Announce Type: new Abstract: Large language models often repeatedly select the same answer even when many alternatives are plausible. Prior work treats this concentration as bias, b
arXiv:2609.09210v1 Announce Type: cross Abstract: Teleoperated demonstrations are often multimodal even when the underlying dynamics are nearly deterministic given the executed action. We argue that t
arXiv:2609.07478v1 Announce Type: new Abstract: Large language models act as strategic agents and models of human choice, yet choosing like a strategic agent does not mean computing like one. We recor
arXiv:2609.06261v1 Announce Type: new Abstract: Vision foundation models trained on Internet-scale RGB datasets enable remarkable capabilities across a range of tasks, from text-to-video generation to
arXiv:2406.04314v4 Announce Type: replace Abstract: Generating visually appealing images is fundamental to modern text-to-image generation models. A potential solution to better aesthetics is direct p
Also note that well aligned models are not sufficient to solve prompt injection by themselves yet. That said, we have solved prompt injection in practice when layering our latest models with prompt in
arXiv:2609.06922v1 Announce Type: new Abstract: Quantization makes vision language models (VLMs) cheaper to store and run by using fewer bits to represent their weights. While unchanged answers on vis
arXiv:2109.08813v1 Announce Type: cross Abstract: Seismic wave velocity of underground rock plays important role in detecting internal structure of the Earth. Rock physics models have long been the fo