Knowledge catalogue
arXiv:2609.37002v1 Announce Type: cross Abstract: High-resolution visual question answering often fails because a multimodal model does not acquire the small, spatially localized evidence needed to an
arXiv:2609.36572v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has been extended to Large Vision-Language Models (LVLMs), and perception-aware methods further en
arXiv:2609.37062v1 Announce Type: cross Abstract: Dynamic layer skipping reduces LLM computation by allowing each token to execute only a subset of the model's layers. However, existing skippers rely
arXiv:2609.37731v1 Announce Type: new Abstract: Activation space and parameter space provide complementary views of model computation. Activations represent information, while weights read, transform,
arXiv:2609.37755v1 Announce Type: new Abstract: Purpose: Most Greek papyri remain unpublished and undigitised; a handwritten text recognition (HTR) pipeline that transcribes them automatically would l
arXiv:2609.36661v1 Announce Type: new Abstract: Visual reprogramming is a parameter-efficient method for adapting pretrained models, yet its training can remain computationally expensive: even with a
arXiv:2609.32530v1 Announce Type: new Abstract: Difference-in-means steering requires activations recorded while a model shows the desired behavior, which a sandbagging model withholds by deliberately
arXiv:2609.34749v1 Announce Type: new Abstract: Real-world driving is inherently multi-agent, yet most existing driving world models generate observations from a single ego vehicle. Independently exte
arXiv:2609.34992v1 Announce Type: new Abstract: Decoding for large language models is typically treated as a collection of isolated sampling strategies, with limited theoretical understanding of the b
arXiv:2609.34807v1 Announce Type: new Abstract: Spatially conditioned diffusion models can embed words and contours in natural-looking images, but vision-language models (VLMs) may fail to recognize t
arXiv:2609.35022v1 Announce Type: new Abstract: The integration of deep learning models into image preprocessing pipelines such as super-resolution introduces a largely unexplored attack vector for ad
arXiv:2609.32705v1 Announce Type: new Abstract: While content-adaptive schemes have delivered notable advances in image super-resolution (SR), existing approaches typically focus on texture complexity
arXiv:2609.33559v1 Announce Type: new Abstract: Recipes such as s1 and LIMO teach a model to reason with little data by showing it the same thousand or fewer worked solutions many times over. Judged w
arXiv:2609.32109v1 Announce Type: new Abstract: We study whether benchmarks reliably guide model selection as developers adapt to evaluation feedback across multiple criteria. We find that the worst-c
arXiv:2609.33407v1 Announce Type: cross Abstract: De novo crystal generation (DNG) models decide where to search in composition space and how to generate structures with one set of weights. We argue t
arXiv:2609.33306v1 Announce Type: new Abstract: Current Transformer-based tracking methods typically stack multiple Transformer blocks with separate parameters to model interactions between the target
arXiv:2609.34262v1 Announce Type: cross Abstract: Agentic benchmarks guide model selection and training. Yet an agent can pass a task without demonstrating the intended capability. Such outcomes const
arXiv:2609.32690v1 Announce Type: new Abstract: Recent multimodal large language models (MLLMs) advance visual reasoning by strengthening both perception and reasoning, implicitly assuming a process t
arXiv:2606.12876v2 Announce Type: replace Abstract: As large language models (LLMs) are increasingly deployed across heterogeneous hardware with varying resource constraints, the ability to adaptively
arXiv:2609.31714v1 Announce Type: new Abstract: Building omni-modal models with physical intelligence requires fine-grained supervision that captures physical evidence such as contact, support, deform
arXiv:2609.31631v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) models enable efficient scaling of large language models but face critical deployment challenges due to massive memory requirem
arXiv:2606.19253v2 Announce Type: replace Abstract: Existing approaches to 3D scene understanding in Vision-Language Models (VLMs) either rely on complex, model-specific geometry encoders or large tra
OpenAI just announced the 200 Pro plan is back but with half the total API spend included, offset by some of their smaller (NOT Astra!) getting a 50% price drop. Does this impact the average user? No.
Ornith-1.5-9B-DFlash pairs the Ornith-1.5-9B model with a DFlash draft model for speculative decoding. https://huggingface.co/ornith-ai/Ornith-1.5-9B-DFlash Ornith-1.5-397B-DFlash pairs the Ornith-1.5
arXiv:2609.33746v1 Announce Type: new Abstract: At each decoding step a language model attends over the key-value (KV) cache of every earlier token, so at long context the attention call is bounded by
arXiv:2609.32259v1 Announce Type: cross Abstract: Recent multi-agent LLM systems increasingly combine heterogeneous models for specialized agent roles. However, text-based communication requires each
Thanks to tcclaviger, vLLM now has expert RAM offloading support (link). This makes frontier models much more accessible on a local setup! I was able to run the original DeepSeek-V4-Flash-Vision-Exp o
arXiv:2609.35249v1 Announce Type: cross Abstract: Pretrained robot manipulation policies such as vision-language-action models (VLAs) or world-action models (WAMs) leave interaction-relevant metric ge
arXiv:2609.35501v1 Announce Type: cross Abstract: Recent agentic symbolic regression approaches increasingly rely on large language models to analyze data, select scientific operations, and refine hyp
arXiv:2609.32672v1 Announce Type: new Abstract: Predicting whether a trained model will generalize under distribution shift remains difficult, especially when target-domain data are unavailable. We in
this is insane, my bonsai 2 27b build is trending on hugging face, first page of text generation, with 20,964 downloads in its first week. it's prismml's ternary bonsai 2 with the qwen 3.8 mtp head gr
arXiv:2609.32824v1 Announce Type: new Abstract: extit{Gromov-Wasserstein Distances} (GWDs) provide quantitative ways of comparing probabilistic distributions defined on different metric spaces by appl
arXiv:2609.31038v1 Announce Type: new Abstract: Time series foundation models (TSFMs) enable cross-domain forecasting, but their development as general-purpose forecasters remains constrained by under
arXiv:2609.31140v1 Announce Type: new Abstract: Most Vision-Language Models (VLMs) are built by extending pretrained Large Language Models (LLMs) with visual modules and multimodal alignment. However,
Claude Sonnet 5.5 New Sonnet model from Anthropic today. They say it 'runs 30%+ faster, and costs up to 30% less for most work' - it's priced the same as Sonnet 5 but appears to beat it on every bench
arXiv:2608.00008v2 Announce Type: replace Abstract: The local deployment of large language models (LLMs) is gaining traction due to privacy concerns and the desire for on-premise inference. However, t
arXiv:2602.09163v2 Announce Type: replace Abstract: Scientific knowledge bases accelerate discovery by curating findings from primary literature into structured, queryable formats for both human resea
arXiv:2609.31386v1 Announce Type: new Abstract: Load-adaptive gravity balancing mechanisms (LA-GBMs) can accommodate various loading conditions by passively changing their characteristics in response
arXiv:2609.31356v1 Announce Type: new Abstract: Vision-Language Models (VLMs) exhibit remarkable zero-shot generalization, yet they often encode unwanted or hazardous stylistic domains such as idealiz
arXiv:2608.25097v2 Announce Type: replace Abstract: Understanding how (multimodal) large language models perform on physics problems requires benchmarks that reflect the difficulty and breadth of expe
arXiv:2512.09172v3 Announce Type: replace-cross Abstract: We tackle continual adaptation of vision-language models to new attributes, objects, and their compositions in Compositional Zero-Shot Learnin
arXiv:2609.30836v1 Announce Type: new Abstract: Deploying small language models (SLMs) on offline, resource-constrained edge devices such as remote sensing satellites presents a fundamental challenge:
arXiv:2606.13020v3 Announce Type: replace Abstract: Three paradigmatic forms of inference recur across scientific reasoning: deduction, induction, and causal abduction. Reliably evaluating LLMs on the
arXiv:2609.30761v1 Announce Type: new Abstract: Most existing human motion generation (HMG) methods use cross-attention modules to inject text semantics, but ignore the importance of bidirectional mod
arXiv:2609.31214v1 Announce Type: new Abstract: Estimating the influence of training examples on model behavior is essential for data debugging, valuation, and attribution. Existing influence estimato
This is brilliant. Recommended reading. It's a great example of how you could potentially build more effective harnesses using System One and System Two models. It explores Jev's potential as a fuzzy
I am running qwen3.8 27b on my dual rtx 3090 (fp8 quant, unquantized cache, 129k context) and I think it works decently well with hermes, opencode etc. But! I am tempted by the new models coming out s
arXiv:2609.29800v1 Announce Type: new Abstract: Adapting multilingual speech foundation models to low-resource languages remains difficult, especially for languages that are poorly represented during
arXiv:2609.28919v1 Announce Type: new Abstract: Harnesses, the products that run AI coding agents, are multiplying, and enterprises are rolling them out to their employees: what started as pilots with
arXiv:2609.29075v1 Announce Type: new Abstract: Cancer registrars, including Oncology Data Specialists (ODSs), must interpret complex and frequently updated coding and staging standards. We developed
arXiv:2609.29848v1 Announce Type: new Abstract: A language model can fail a syntactic test in two distinct ways: by not encoding the relevant structure, or by encoding it but failing to use it at the
arXiv:2609.30144v1 Announce Type: new Abstract: Most benchmarks hand the model a question. EnigmaForge hands it a stack of old documents and no question at all. Buried in the letters, receipts, and lo
arXiv:2609.28741v1 Announce Type: new Abstract: Multi-modal multitasking models have shown strong performance on remote sensing datasets. However, because these models are trained on heterogeneous dat
arXiv:2604.01754v2 Announce Type: replace-cross Abstract: Mathematical reasoning is a hallmark of human intelligence, and whether large language models (LLMs) can meaningfully perform it remains a cen
arXiv:2603.26797v2 Announce Type: replace Abstract: Large language models are increasingly used to generate financial alpha signals, but many have memorized the historical data in their training corpo
arXiv:2603.04756v3 Announce Type: replace Abstract: MOOSEnger is a modeling and simulation AI agent framework for the Multiphysics Object-Oriented Simulation Environment (MOOSE) ecosystem, built aroun
arXiv:2609.28587v1 Announce Type: cross Abstract: Large language models can interpret natural lan- guage, yet robust decisions remain challenging. Jev-like models expose structured choices, but these
arXiv:2609.29251v1 Announce Type: new Abstract: CAR-bench evaluates whether tool-using agents stay reliable under real-world uncertainty, executing every tool inside the evaluator so that each tool-re
arXiv:2609.28693v1 Announce Type: new Abstract: Large Language Model (LLM) agents struggle to scale safely when exposed to vast enterprise toolsets. Providing an agent with access to every internal to
arXiv:2609.26176v2 Announce Type: replace-cross Abstract: Encoded-prompt attacks are evaluated almost entirely on their harmful arm: a benchmark sends obfuscated harmful requests and reports how often