Search: “benchmark”

Exploring the Underwater World Segmentation without Extra Training

arXiv:2511.07923v3 Announce Type: replace Abstract: Accurate segmentation of marine organisms is vital for biodiversity monitoring and ecological assessment, yet existing datasets and models remain la

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

FoveaTree: Information-Density Trees for Query-Guided Visual Token Pruning in Video Large Language Models

DGX agent

arXiv:2610.16744v1 Announce Type: new Abstract: Video large language models (Video LLMs) encode each frame into hundreds of visual tokens, making video inference expensive. Existing methods reduce thi

model-releasesarxiv-cs-cv

14 Oct 2026

FunDM: Cheap & Label-Free Decision Models from Frozen Language Models

arXiv:2610.17175v1 Announce Type: new Abstract: Decision models such as Jev score candidate answers or actions. A small decision head trained label-free can recover about 99% of a frozen language mode

GA-SVD: Generation-aware Singular Value Decomposition for LLM Compression

arXiv:2610.15398v1 Announce Type: new Abstract: Large language models (LLMs) demonstrate strong generalization across a broad range of reasoning and generation tasks, but their large parameter counts

Genesis: Visual Augmentation Primitives for Multimodal Chain-of-Thought Reasoning

arXiv:2610.15565v1 Announce Type: new Abstract: Recent multimodal large language models achieve promising results on visual reasoning tasks, approaching human-level performance on existing STEM benchm

Historical Anchors for Scientific Forecasting on Human-Aware Knowledge Graphs

arXiv:2610.14023v1 Announce Type: new Abstract: Predicting which scientific connections will form next can point researchers toward the most promising leads. One successful approach learns representat

History, Adaptation, and Hidden Gain: Two Prospective Tests of Learned Predictive Control in a Synthetic Neural Population

arXiv:2610.14368v1 Announce Type: new Abstract: We study learned one-step predictive control of a synthetic population of 64 leaky integrate-and-fire neurons whose light sensitivity is scaled by a hid

HYDRA: Representation Harmonized Tokenization for Multimodal Generation and Understanding

arXiv:2603.15228v3 Announce Type: replace Abstract: Unified Multimodal Models struggle to bridge the fundamental gap between the abstract representations needed for visual understanding and the detail

InfraWorld: Motion-Aware World Modeling for Roadside 4D Occupancy Forecasting

arXiv:2610.14544v1 Announce Type: new Abstract: Roadside 4D occupancy forecasting predicts future semantic occupancy in a fixed infrastructure coordinate system. Roadside scenes are globally structure

KARAMEL: Knowledge-Aware Ranking for Multimodal Entity Linking in the Arts via Heterogeneous Graphs

arXiv:2610.16502v1 Announce Type: new Abstract: Discovering and disambiguating entities from multimodal and heterogeneous data is a key challenge in discovery science, where relevant knowledge is ofte

Learning Operators for Symmetry Discovery: From Lie Groups to Local Symmetry Geometry

arXiv:2610.16884v1 Announce Type: new Abstract: Continuous symmetry discovery is typically formulated as a separate estimation problem for each function or dataset. We instead investigate whether the

researcharxiv-cs-lg

Look Back to Move Forward: Learning to Revisit for Agentic Compositional Image Generation

arXiv:2610.16224v1 Announce Type: new Abstract: Text-to-image models have made remarkable progress in visual fidelity and instruction following, yet complex compositional generation remains challengin

Match the Canvas, Not Just the Answer: Predictability-Aware SFT for Diffusion Language Models

arXiv:2610.18001v1 Announce Type: new Abstract: Diffusion language models (DLMs) enable flexible parallel generation by resolving tokens according to the model's evolving predictability rather than a

Merit: Making every modality earn its contribution in multimodal attention decoding

arXiv:2610.17507v1 Announce Type: new Abstract: More than 1.5 billion people have hearing loss. When several people talk at once, hearing aids amplify every voice alike. Auditory attention decoding (A

MiMaAlign: Hierarchical Micro-Macro Alignment for Physically Plausible Video Generation

arXiv:2610.16482v1 Announce Type: new Abstract: Text-to-video (T2V) diffusion models have made remarkable progress in generating high-fidelity visuals, yet their outputs still often exhibit physically

Online LLM Personalization from Pairwise Preferences Across Users

arXiv:2610.17203v1 Announce Type: new Abstract: Large language models can answer the same question in different styles, but users' preferred styles are hidden and vary across users and questions. We s

Persistence Positional Encoding: A Minimal Transformer Recipe for Persistence Diagrams

arXiv:2610.17848v1 Announce Type: new Abstract: Persistence diagrams (PDs) represent multiscale topological information as multisets of birth-death pairs (b,d). How best to use them with Transformers,

researcharxiv-cs-lg

Physical SLAM: Physics-aware Simultaneous Localisation and Mapping

model-releasesarxiv-cs-ro

arXiv:2610.17260v1 Announce Type: new Abstract: To reconstruct a predictive 4D representation of the physical world from monocular video captured by a moving camera, we jointly reconstruct dynamic Gau

ProbabilityBench: Model Evaluation and Calibration with Exact Ground-Truth Uncertainty

arXiv:2610.17016v1 Announce Type: new Abstract: Evaluating model uncertainty on real-world data is challenging because each instance provides only a single observed label, leaving true conditional cla

Query-Adaptive Neuron Steering for LLMs

arXiv:2610.14519v1 Announce Type: new Abstract: Effectively eliciting the capabilities of large language models (LLMs) at inference time remains an important challenge. Existing studies address this c

RadarGSOcc: Doppler-Aware 3D Semantic Occupancy Prediction Using 4D Radar Tensor with Geometric-Semantic Feature Extraction

arXiv:2610.18686v1 Announce Type: new Abstract: 3D semantic occupancy prediction is crucial for autonomous driving, but camera-based methods degrade in low-light conditions and LiDAR-based methods are

Representation Drift in Continual Learning: LoRA versus N-gram Memory

arXiv:2610.14921v1 Announce Type: new Abstract: A language model that keeps learning online must not slowly erode its own base capabilities. We ask which part of a modern LLM should absorb continual u

Rescuing Points Trapped Inside Incorrect Decision Regions: A target for Evaluating and Improving Forward-Forward Learning

arXiv:2610.16347v1 Announce Type: new Abstract: Forward-Forward learning replaces end-to-end backpropagation with local, layer-wise training objectives. Its variants are evaluated primarily by classif

researcharxiv-cs-lg

RSIgen: Reciprocal Self-Improvement for Agentic Visual Generation

arXiv:2610.18688v1 Announce Type: new Abstract: Agentic visual generation produces experience that can improve the native understanding (agent) and generation roles of a unified multimodal backbone, b

Sabi-Noema: A capability and efficiency focused fine-tune of Qwen 3.8 27B

model-releasesr-localllama

tl;dr: new qwen finetune that simultaneously pushes the frontiers of small model capability and reasoning efficiency. Hey everyone, We publish two post-trained versions of Qwen 3.8 27B called Sabi-Noe

SCOPE Before You Sample: Conformal Scores Guide Posterior Exploration for Fast and Calibrated Bayesian Predictive Inference

arXiv:2610.17146v1 Announce Type: cross Abstract: Bayesian predictive regions account for parameter uncertainty but can be poorly calibrated under model misspecification, even with expensive full MCMC

SLEM: Spectrum-Aware Latent Expert Merging

arXiv:2610.16484v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) models scale model capacity efficiently by activating only a small subset of experts for each input, while still requiring all

Squeeze-Aware-YOLO: Early Channel Recalibration and Multi-Scale Fusion for Real-Time Traffic Sign Detection

arXiv:2610.15604v1 Announce Type: new Abstract: Traffic sign detection plays a vital role in intelligent transportation systems, supporting the safe operation of advanced driver assistance systems and

Truncation Samplers Are Nested: Comparing Decoding Rules Is a Measurement Problem

arXiv:2610.15570v1 Announce Type: cross Abstract: Settling one comparison between two truncation samplers has cost six thousand A100-hours, and nothing says in advance how far apart two samplers can b

TwinPrune: Density-Aware Two-Phase Token Pruning for Vision-Language Models

arXiv:2610.17530v1 Announce Type: new Abstract: We propose a unified spatial-density signal to score token importance for both inference phases: bipartite merge for prefill and KV eviction for decode,

researcharxiv-cs-cv

Safety

Variance-Optimal State Resampling for Reinforcement Learning with Verifiable Rewards

arXiv:2610.15888v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards spends much of its rollout budget on states that yield no learning signal. We recast this as a question a

safetyarxiv-cs-lg

VideoJev: System One Model for Video Understanding and Interaction

arXiv:2610.18615v1 Announce Type: new Abstract: Visual agents must turn video observations into timely decisions to track progress and choose what to do next. Building on the emerging System One parad

researcharxiv-cs-cv

VIEWCRAFT: Evolving What Multimodal Agents See and How They Read It

arXiv:2610.15915v1 Announce Type: new Abstract: Reusable skills improve multimodal agents without updating model weights, but their effectiveness depends on both the evidence presented and the procedu

What Limits Cross-Task Generalization in World Action Models?

arXiv:2610.16641v1 Announce Type: new Abstract: Cross-task generalization remains difficult for world action models (WAMs), even when their predicted futures depict plausible manipulation. This gap ca

researcharxiv-cs-ro

YOUDrive: Driving Style as a Composable Direction for End-to-End Autonomous Driving

arXiv:2610.14802v1 Announce Type: new Abstract: End-to-end driving policies can achieve strong task performance while offering limited control over their driving style. Driving style describes how a v

zPocket: Better Pocket Finding with Frozen Co-Folding Features

arXiv:2610.17696v1 Announce Type: new Abstract: Identifying protein-ligand binding sites (pockets) is important for understanding protein function and developing drugs. However, no public method has l

I built a pytorch library of pretrained models for 3D point clouds following timm standards [P]

model-releasesr-machinelearning

https://preview.redd.it/m5e9anluo8vh1.png?width=1080&format=png&auto=webp&s=4c80687e6fe40036843a97ed8975ce46eaab0662 I am excited to share a project I have been working on during the past few years du

13 Oct 2026

llm-gemini 0.35

model-releasessimon-willison

Release: llm-gemini 0.35 I added support to llm-gemini for the new Gemini 4 Argon model, made available today via the Gemini API and for customers of their 100/month subscription plan. Here are the lo

13 Oct 2026

Using Gemini 4 makes you realize how many other models must be distilling Claude/GPT (or using a similar training pipeline). It feels quite …

model-releasesethan-mollick--x

On October 13 2026, researcher Ethan Mollick tweeted that observing Gemini 4’s behavior reveals how many other LLMs are likely distilled from Claude or GPT—and thus use a similar training pipeline—mak

13 Oct 2026

A Record of the Process of Increasing MFU from 21% to 50% While Pre-training 21B MoE Using a Single GH200 Chip (Includes Code and Logs)

model-releasesr-localllama

I am an independent researcher. Due to budget constraints, I shifted my 21B MoE pre-training—which I had originally planned to run on 64 H100s—to a single NVIDIA GH200 ($2.29 per hour), and I have com

Amortized Counterfactual Prediction with Graph Neural Networks

arXiv:2610.13724v1 Announce Type: cross Abstract: Counterfactual inference asks how an observed outcome might have differed under different circumstances, such as whether a patient would have recovere

APEX-Harm: Evaluating Agent Safety in Shared Professional Workspaces

model-releasesarxiv-cs-cl

arXiv:2610.13623v1 Announce Type: cross Abstract: Professional agents increasingly work through command-line interfaces and multiple tools over shared workspaces containing files, messages, and busine

Beyond Temporal Proximity: Explanatory World Models for Credit Assignment in Reinforcement Learning

arXiv:2610.13504v1 Announce Type: cross Abstract: Generative Recommenders increasingly use Reinforcement Learning (RL) to optimize outcomes that may appear hundreds or thousands of interactions after

Local Ai

Beyond the Final Embedding: Retrieving Memories from Token-Level Evidence

arXiv:2610.13277v1 Announce Type: cross Abstract: Memory-augmented assistants retrieve evidence from stored conversations and documents to answer queries. Most embedding-based retrievers compare singl

local-aiarxiv-cs-cl

Bistable by Construction: Wall-Clock-Calibrated State Monitors Have No Moment-Detection Regime at Agent Cadence

arXiv:2606.19386v2 Announce Type: replace-cross Abstract: Runtime monitors for autonomous agents commonly threshold an accumulated internal state - a behavioural baseline, a drift statistic, or, in ou

Constructing Pareto Frontiers of Optimal Reasoning Termination Methods for Overthinking in Reasoning Models

arXiv:2610.13570v1 Announce Type: new Abstract: Reasoning models can benefit from test-time scaling, but continued reasoning may become ineffective or harmful, a phenomenon termed overthinking. Evalua

Data-driven Optimal Sensor Placement for Aerosol Monitoring in the UK

arXiv:2610.12658v1 Announce Type: new Abstract: We develop a data-driven sensor-placement framework and apply it to aerosol monitoring in the United Kingdom (UK). A trained convolutional Gaussian neur

Deploy decision models with llama.cpp on Amazon SageMaker AI

model-releasesaws-ml-blog

Learn how to deploy decision models with llama.cpp on Amazon SageMaker AI. This technical how-to serves OpenAI-compatible chat completions with the AWS Deep Learning Container for llama.cpp, then exte