Search: “benchmark”

arXiv:2609.35469v1 Announce Type: cross Abstract: Autoregressive Vision-Language-Action (VLA) models offer a scalable path to robot learning, yet existing action tokenizers treat tokenization as a com

applicationsarxiv-cs-lg

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

Rethinking Circuit Evaluation: Do Circuits Explain Model Errors?

DGX agent

arXiv:2609.35686v1 Announce Type: new Abstract: Mechanistic interpretability (MI) aims to explain a model's behaviour through analyzing its internal computations; circuit-based explanations aim to iso

model-releasesarxiv-cs-lg

29 Sept 2026

Rethinking Visual Token Compression for Video Large Language Models: A Simple Yet Strong Baseline

arXiv:2609.35394v1 Announce Type: new Abstract: Video Large Language Models (Video LLMs) have achieved remarkable progress in video understanding, but their inference efficiency is constrained by the

Agents

Retrieved but Not Delivered: Multimodal Memory Delivery for Long-Term Agents

arXiv:2609.32590v1 Announce Type: new Abstract: Work on memory for multimodal agents optimizes what is written, updated and retrieved. Between retrieval and the answer, however, is a stage that multim

agentsarxiv-cs-cv

Reuse or Relearn? A Spectral View of Earth Observation Foundation Models

arXiv:2609.32756v1 Announce Type: new Abstract: Foundation models are rarely used as generic, frozen feature extractors; instead, they are fine-tuned for the target downstream application. This practi

Local Ai

Revisit to Segment: Working Memory Distillation for Reasoning Segmentation

arXiv:2609.34863v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) have approached image segmentation by reasoning about visual content and predicting target locations. Their gen

local-aiarxiv-cs-cv

Revisiting Diffusion Fine-Tuning for Unsupervised Domain Adaptation

arXiv:2609.33716v1 Announce Type: new Abstract: Diffusion-based unsupervised domain adaptation (UDA) improves cross-domain transfer by generating target-specific synthetic data for downstream adaptati

Tutorials

Revitalizing Medical Time Series with Vision-Informed Retrieval: A Vision-Language Perspective

arXiv:2609.34652v1 Announce Type: new Abstract: Medical time series (MedTS) underpin many clinical classification tasks, yet existing methods usually represent them only as numerical sequences and und

tutorialsarxiv-cs-cv

RewardExplainer: Learning Reward Model Explanations from Counterfactual Preference Feedback

arXiv:2609.33989v1 Announce Type: cross Abstract: Reward models (RMs) are a key component of large language model post-training, providing reward signals for subsequent reinforcement learning. However

RIPE-MambaSpike: Resolution-Independent Spiking-State-Space Interfaces for Parameter-Efficient Event-Based Vision

arXiv:2609.32537v1 Announce Type: new Abstract: Spiking-Mamba hybrids reach strong accuracy on event-based vision, but existing designs often require tens of millions of parameters. Much of that cost

RoboFFT: Finetuning generative robot policy via online reinforcement learning with forward process

arXiv:2609.32236v1 Announce Type: new Abstract: Generative models, such as diffusion and flow-based models, have shown strong promise for robot policy learning by capturing complex and multimodal acti

safetyarxiv-cs-ro

RoPE is Dead, Long Live RoPE: Towards Scalable Data-aware Positional Encodings

arXiv:2609.34556v1 Announce Type: cross Abstract: Transformers process tokens without any inherent notion of order, making positional encoding a fundamental requirement rather than an architectural re

Routing Without Embeddings: Fast And Interpretable Routing With Regular Expressions

arXiv:2609.34326v1 Announce Type: new Abstract: Large Language Model (LLM) routers commonly rely on neural query embeddings, with larger encoders expected to better capture query intent and difficulty

RubiCap: Rubric-Guided Reinforcement Learning for Dense Image Captioning

arXiv:2603.09160v2 Announce Type: replace Abstract: Dense image captioning is critical for vision-language pretraining and for text-to-image generation, but scaling expert-quality annotations is prohi

safetyarxiv-cs-cv

Safe Score Matching: Diffusion Policies with Hamilton-Jacobi Reachability for Online Safe Reinforcement Learning

arXiv:2609.33337v1 Announce Type: new Abstract: Online safe reinforcement learning (RL) seeks policies that maximize reward while satisfying safety constraints. A popular line of research in safe RL r

SafeMol: Dual-Modality Safety Alignment for Molecular Multimodal Models

arXiv:2609.33640v1 Announce Type: new Abstract: Molecular multimodal models support diverse understanding and generation tasks but may introduce safety vulnerabilities when handling hazardous molecule

SAMBAR: Selective Anchoring via Method of Multipliers for Balanced Knowledge Acquisition and Retention in Vision-Language-Action Models

arXiv:2609.32108v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models leverage large-scale pretraining to ultimately achieve generalist manipulation. Deployed VLA policies must support c

Scalable In-Domain Self-Supervised Foundation Model for Dense Representation Transfer in High-Resolution Plant Imaging

arXiv:2609.32183v1 Announce Type: new Abstract: High-resolution plant imaging enables detailed characterization of plant morphology, but dense scientific analysis remains limited by costly pixel-level

SciGen-Verifier: A Multimodal Reasoner for Explainable Verification in Scientific Image Generation

arXiv:2609.33399v1 Announce Type: new Abstract: In realistic education, a solution is often expressed not only in words but in a drawing--a circuit, a geometric construction, a function plot--and a te

Selecting Diverse SFT Traces Improves Post-RL Generalization

arXiv:2609.33780v1 Announce Type: new Abstract: Verified solutions are not equally useful for preparing reasoning models for reinforcement learning (RL). We present a comprehensive study of route dive

Self-Evolving Time-Series Forecasting Agents with Episodic Memory and Online Policy Learning

arXiv:2609.32689v1 Announce Type: new Abstract: LLM-based agents are increasingly used for time-series forecasting because they can organise contextual information, perform multi-step analysis, and gu

SignFLIP: A Unified Model for Sign Language Translation and Generation via Stage-wise Alignment at Scale

arXiv:2609.35225v1 Announce Type: cross Abstract: Sign language translation and generation share the goal of bidirectional alignment between text and sign representations. However, existing approaches

safetyarxiv-cs-cv

Silent Failures in Agentic Security Evaluation: A Validated Harness for Tool-Call Mediation Under Indirect Prompt Injection

arXiv:2609.32691v1 Announce Type: cross Abstract: LLM agents that invoke privileged tools are vulnerable to indirect prompt injection (IPI), in which adversarial instructions embedded in retrieved dat

Simple Diffusion Language Models Are More Effective Few-Step Generators Than Reported

arXiv:2609.33947v1 Announce Type: cross Abstract: Diffusion language models (DLMs) promise fast parallel generation, yet high-quality samples often require large number of refinement steps, which dimi

Applications

Simple Extensions of Single-Objective Acquisition Functions and Hedge Strategies for Multi-Objective Bayesian Optimization

arXiv:2609.31940v1 Announce Type: new Abstract: Multi-objective Bayesian optimization (MOBO) is commonly approached through specialized acquisition functions or scalarization schemes designed to expli

applicationsarxiv-cs-lg

Simulation-Free Learning of GP-SDEs from Irregular Observations

arXiv:2609.33112v1 Announce Type: new Abstract: Gaussian process stochastic differential equations (GP-SDEs) provide a flexible Bayesian model for unknown continuous-time state dynamics with uncertain

SketchSSM: Write to the Full State, Read from a Compact Sketch

arXiv:2609.33051v1 Announce Type: new Abstract: Hybrid-attention models replace most softmax attention layers with linear attention, reducing KV-cache growth and enabling larger decode batches where r

SkillPE: Creativity-Oriented Cinematic Skill Evolution for Text-to-Video Prompt Engineering

arXiv:2609.34335v1 Announce Type: new Abstract: Achieving high-quality, cinematic results in text-to-video generation remains challenging for non-experts, whose prompts often lack professional narrati

SlimWise: Decoupling Expert Pruning Across Prefill and Decode for Efficient MoE Serving

arXiv:2609.34117v1 Announce Type: new Abstract: Mixture-of-experts (MoE) models activate few experts per token, yet batched decoding can access nearly the entire expert pool, making expert-weight traf

SLIP-VLA: Single-Step Latent Imagination for Policy Learning in Vision-Language-Action Models

arXiv:2609.33575v1 Announce Type: new Abstract: Vision-Language-Action models are increasingly effective for robotic manipulation, yet most predict actions directly from current observations without e

safetyarxiv-cs-ro

Smoothing the Semantic Landscape: Generalizable AI-Generated Image Detection via Text-Induced Flatness

arXiv:2603.00717v3 Announce Type: replace Abstract: Recent advances in generative models have exposed a fundamental weakness of image fake detectors: they often fail to generalize to unseen synthesis

Agents

SOR-Nav: Search or Relocate? Context-Gated Exploration and Cross-Region Relocation for Object Navigation

arXiv:2609.34707v1 Announce Type: new Abstract: Object navigation requires an embodied agent to find an object in an unseen environment under partial observability and a limited motion budget. Existin

agentsarxiv-cs-ro

Spectral Reversal: Counteracting Singular Value Bias for Graph Prompting

arXiv:2609.32143v1 Announce Type: new Abstract: Pre-training Graph Neural Networks (GNNs) via self-supervised learning has become a dominant paradigm, yet efficiently adapting frozen encoders remains

Tutorials

SphMind: Towards Robust, Training-Free VLM-based Spatial Reasoning with a 360 Camera

arXiv:2609.33462v1 Announce Type: new Abstract: Omnidirectional or 360 cameras provide embodied AI agents with a holistic, wide field-of-view (FoV) view of their surroundings, motivating the use of Mu

tutorialsarxiv-cs-cv

SpikeLite: Lightweight Spiking Neural Networks for Time-Series Forecasting

arXiv:2609.35097v1 Announce Type: new Abstract: Spiking neural networks (SNNs) offer an energy-efficient paradigm for time-series forecasting through spike-driven computation. However, recent SNN fore

Agents

Sprout: Building Dynamic Memory While Reasoning for Agentic Video Understanding

arXiv:2609.35497v1 Announce Type: new Abstract: Long video understanding relies on video memory to overcome the context limits of multimodal large language models. Existing methods follow a build-then

agentsarxiv-cs-cv

SR4-Fit: A Unified Interpretable Rule-Based Machine Learning Framework for Informative and Trustworthy Decision-Making

arXiv:2609.34019v1 Announce Type: new Abstract: In many high-stakes applications, machine learning is dominated by black-box models that require post hoc explanations to justify their predictions. The

Applications

STAR: Adaptive Spatial-Temporal Normalization for Unified Microservice Incident Management

arXiv:2609.31645v1 Announce Type: new Abstract: Automated incident management in large-scale microservice systems relies on learning robust representations from multimodal observability data, includin

applicationsarxiv-cs-lg

StarOR: Synergizing Tree Search and Test-Time Reinforcement Learning for Optimization Modeling

arXiv:2606.15197v3 Announce Type: replace Abstract: Optimization modeling is inherently hierarchical, requiring a precise sequence of symbolic commitments. Traditional learning-based automated optimiz

STeP: Signal Temporal Logic for Precise Specifications for Action Generation with Vision Language Models

model-releasesarxiv-cs-ro

arXiv:2607.18580v2 Announce Type: replace Abstract: Natural-language robot instructions often specify more than a coarse task goal: they may impose spatial, temporal, and logical requirements that mus

StoryEngine: A State-Grounded Agentic Framework for Video Storytelling

arXiv:2609.33627v1 Announce Type: new Abstract: Despite recent progress in agentic multi-shot video generation, producing coherent and consistent long-form stories remains challenging. Existing agenti

STRIDE: State-Transition Representation via Increment Dynamics and Evolution

arXiv:2609.32367v1 Announce Type: new Abstract: We introduce STRIDE (State-Transition Representation via Increment Dynamics and Evolution), which defines states through derivative fingerprints and lea

Structure-First Point Cloud Learning with Mapper Region Graphs

arXiv:2602.05522v2 Announce Type: replace Abstract: Robust 3D point cloud classification is often pursued by scaling up backbones or relying on specialized data augmentation. We instead ask whether st

Structured Latent Modeling for Supervised Multimodal Information Decomposition

arXiv:2609.35502v1 Announce Type: new Abstract: Multimodal prediction relies on diverse forms of evidence: information repeated across modalities, cues specific to a single source, and complex cross-m

Structured Neural SDEs for Functional Calibration

arXiv:2609.34831v1 Announce Type: new Abstract: Neural Stochastic Differential Equations (Neural SDEs) provide flexible continuous-time generative models, but generic neural drift and diffusion networ

SubRot: Signed Gradient Subspace Calibration for VLM Rotation Quantization

arXiv:2609.34884v1 Announce Type: new Abstract: Post-training quantization reduces the deployment cost of vision-language models (VLMs), but preserving multimodal capabilities at low bit widths remain

SymDrift: One-Shot Generative Modeling under Symmetries

arXiv:2605.06140v2 Announce Type: replace Abstract: Generative modeling of physical systems, such as molecules, requires learning distributions that are invariant under global symmetries, such as rota

Synthetic data for ratemaking: imputation-based methods vs adversarial networks and autoencoders

arXiv:2509.02171v3 Announce Type: replace-cross Abstract: Actuarial ratemaking depends on high-quality data, yet access to such data is often limited by the cost of obtaining new data, privacy concern