Search: “models”

Every Ablation Is a Dose: Counterweights and the Semblance of Self-Repair

arXiv:2610.02173v1 Announce Type: new Abstract: Ablate a component of a language model, and other components often appear to adjust and compensate. This phenomenon, termed self-repair, has been observ

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Model Releases

FedLore: Communication and Memory Efficient Federated Learning via Shared Gradient Low-Rank Projection

DGX agent

arXiv:2610.01620v1 Announce Type: new Abstract: Federated training of foundation models is constrained by client memory and communication costs. LoRA-based methods reduce these costs through low-rank

model-releasesarxiv-cs-ai

2 Oct 2026

Local Ai

From Task Mixtures to Specialized Experts

arXiv:2610.00580v1 Announce Type: new Abstract: In collaborative foundation model fine-tuning, client data is rarely homogeneous. Instead, clients typically possess unknown mixtures of distinct data d

local-aiarxiv-cs-lg

GPEC: Efficient Pre-LLM Gaussian Process Embedding Correction for Cardiac Video Caption Generation

arXiv:2610.00196v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) have shown strong potential for video understanding and caption generation, but their performance may decline i

GUI-HARVEST: Self-Improving GUI Agents through Evidence-Driven Harness Evolution

arXiv:2610.00948v1 Announce Type: cross Abstract: The executable harness surrounding a GUI model determines how observations are assembled, actions are executed, and verification, recovery, and termin

In Dialogue with Intelligence: Toward Insightful Co-Augmentation

arXiv:2505.22767v4 Announce Type: replace-cross Abstract: Dialogue with a large language model can lead a person to insight: a sudden change in how they understand a problem. This perspective asks how

IrekoGPT: Turning Structured Pruning into Post-Hoc Slimmable LLMs

arXiv:2610.00426v1 Announce Type: new Abstract: We introduce IrekoGPT, a post-hoc method for converting pretrained LLMs into slimmable models whose width can be adjusted at inference time. Building on

JevSpawn: Adaptive Agentic Inference through Compositional Action Spaces

arXiv:2610.00437v1 Announce Type: new Abstract: LLM agents generate intermediate reasoning and actions token by token, making extended interactions slow and computationally expensive. Jev-style models

Learning Rate Transfer for Hybrid Transformer-SSM Architectures

arXiv:2610.01172v1 Announce Type: new Abstract: We study learning rate (LR) scaling for hybrid architectures combining Transformer and State-Space Model (SSM) blocks, a class adopted by several recent

M^2Weather: A Benchmark for Joint Multi-Station and Multi-Variable Weather Forecasting

arXiv:2610.00370v1 Announce Type: new Abstract: Station weather forecasting is fundamentally shaped by both complex spatial dependencies across stations and strong physical coupling among weather vari

Mapping the RAG Landscape: A Four Axis Taxonomy of Efficiency, Defense, Interactivity, and Reasoning

arXiv:2610.01936v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated remarkable fluency across many tasks but remain limited by their static, parameter bound knowledge and th

Mem++: Non-Destructive Memory for Long-Term Organizational LLM Agents

arXiv:2610.02002v1 Announce Type: cross Abstract: Large Language Model (LLM) agents now take part in organizational work, where many authors record decisions across documents over months. Because a re

MMMG: a Comprehensive and Reliable Benchmark for Multitask Multimodal Generation

arXiv:2505.17613v2 Announce Type: replace Abstract: Automatically evaluating multimodal generation presents a significant challenge, as automated metrics often struggle to align with human evaluation,

MoRA: MoE Pruning via Router Bias Learning and Expert Approximation

arXiv:2610.00367v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) models enable parameter scaling with limited per-token computation by activating only a small subset of experts for each token,

Multi-Reference Path Tracking Control for an Agricultural Tractor with Nonlinear Model Predictive Control

arXiv:2610.00057v1 Announce Type: new Abstract: Guiding a tractor along a predefined reference path is a key component of precision agriculture. This study develops a path tracking controller based on

researcharxiv-cs-ro

Not Too Hard, Not Too Easy: Learning from Intermediate States for LLM Structured Reasoning

arXiv:2609.33149v2 Announce Type: replace Abstract: A common principle of effective learning is to practice material that is neither already mastered nor too difficult to permit progress. We ask how t

Reasoning as Pattern Matching: Shared Mechanisms in Human and LLM Everyday Reasoning

arXiv:2606.13607v3 Announce Type: replace Abstract: When large language models (LLMs) fail to generalize or make content-sensitive errors in reasoning, it is often taken as evidence that LLMs are not

Target-Dependent Limits of Causal Repair: A Leading-Log Frontier in a Gaussian Model

arXiv:2610.00424v1 Announce Type: cross Abstract: Knowing how much a causal predictor could improve need not reveal the gain of the repair actually learned. We quantify this gap in a scalar Gaussian c

The Delegation Danger Band: Why Mid-Capability Sub-Agents Over-Trust Inherited Stale State

arXiv:2610.00041v1 Announce Type: cross Abstract: Agent frameworks increasingly delegate work by forking sub-agents; a common default makes the child inherit the parent's full working context. We meas

Safety

Towards a General Humanoid Loco-Manipulation Model via Egocentric Whole-Body Human Data Pretraining

arXiv:2610.00438v1 Announce Type: new Abstract: Humanoid whole-body manipulation has advanced rapidly, enabling policies to coordinate locomotion, posture, bimanual interaction, and dexterous hand mov

safetyarxiv-cs-ro

Two Clocks in Diffusion MLLMs: When Answers Stabilize Before Rationales Unfold

arXiv:2610.00953v1 Announce Type: cross Abstract: An answer candidate in a masked diffusion MLLM can stabilize while its rationale is still unfolding. We distinguish retrospective stabilization of the

A helps B while B hurts A: directed transfer in instruction-tuning mixture

arXiv:2609.39702v1 Announce Type: new Abstract: Adapting a language model to a specialized corpus means choosing which instruction-tuning tasks to train on under a fixed budget, and testing one choice

Aligning Thoughts with Answers: Probability Rewards to Tame Thinking Drift

arXiv:2609.39183v1 Announce Type: new Abstract: This paper studies extbf{thinking--answer consistency} in vision-language models. We focus on Visual Intention Grounding, where a model infers a target

researcharxiv-cs-cv

Certification-Based Differentially Private Learning

arXiv:2609.39629v1 Announce Type: new Abstract: Differential privacy (DP) in machine learning is typically achieved by adding noise to model parameters (private learning) or to model outputs (private

Safety

Code to Control: Synthesizing Parameterized Reactive Controllers

arXiv:2609.38733v1 Announce Type: new Abstract: Recent LLM-based approaches to control either invoke a language model to select actions or synthesize world models that require planning at every decisi

safetyarxiv-cs-ai

Tutorials

Coverage Before Control: Route-Instruction Grounding and Steering for Controllable Retrosynthesis

arXiv:2609.39955v1 Announce Type: new Abstract: Single-step retrosynthesis models are commonly evaluated by their ability to recover recorded reactions. In practice, chemists may need to choose among

tutorialsarxiv-cs-ai

EvoDuet: Bilevel Co-Evolution of Web Searching and Task Solving for Scientific Discovery

model-releasesarxiv-cs-cl

arXiv:2609.40340v1 Announce Type: new Abstract: Evolutionary search with large language models (LLMs) can stall when progress requires external knowledge the model lacks. Supplying relevant documents

Safety

EVOKE: Eliciting World Knowledge in Agents for Transferable Decision-Making

arXiv:2609.38334v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed as agents for multi-step decision-making, yet transfer poorly to unseen environments. World-model

safetyarxiv-cs-cl

Faithful Dual-constrained Erasure for Robust LLM Safety Alignment

arXiv:2609.39279v1 Announce Type: cross Abstract: Machine unlearning has emerged as a crucial mechanism for removing hazardous knowledge and enforcing safety alignment in Large Language Models (LLMs).

GroundAnything: Reconciling Parallel Decoding with Precise Visual Grounding at Flash Speed

arXiv:2609.39600v1 Announce Type: cross Abstract: Autoregressive (AR) grounding models serialize spatial predictions, introducing sequential latency and imposing a causal order on output tokens. We vi

Tutorials

Hermes: Learning Contextual Reasoning Unlocks Test-Time Scaling

arXiv:2609.38332v1 Announce Type: cross Abstract: Test-time scaling improves model performance by allocating additional compute during inference. Using this compute effectively across multiple context

tutorialsarxiv-cs-ai

How Much Is an AI Token Worth? Scaling Laws for Wild AI-Generated Web Text

arXiv:2609.40295v1 Announce Type: new Abstract: Web text makes up the majority of pretraining data and is increasingly AI-generated. After applying FineWeb quality filtering, we find that 27.5% of tok

researcharxiv-cs-cl

Inductive Visual Logic for Few-Shot Out-Of-Distribution Adaptation in VLMs

arXiv:2609.38362v1 Announce Type: new Abstract: Generative vision-language models (VLMs) such as Qwen-VL and LLaVA achieve strong zero-shot performance on tasks overlapping with their pretraining dist

KilometerVision: A New Frontier for Large-Scale Spatial Intelligence in VLMs

arXiv:2609.39588v1 Announce Type: new Abstract: We push the frontier of large-scale spatial intelligence in Vision-Language Models (VLMs) and introduce the first benchmark that probes geographical lay

Lasting Effects of Abstract Pretraining Beyond Perplexity

arXiv:2609.38764v1 Announce Type: new Abstract: Language models are typically pretrained from random initialization. Recent work challenges this convention, showing that a brief warm-up on abstract, a

researcharxiv-cs-lg

Replay on Demand: An Emergent Curriculum for Balancing Adaptation and Forgetting in Continued Pretraining

arXiv:2609.40089v1 Announce Type: new Abstract: Continued pretraining enables language models to adapt to new domains and knowledge, but often at the cost of forgetting previously acquired capabilitie

researcharxiv-cs-lg

Representation Risk in Pretrained Image Encoders

arXiv:2609.35470v1 Announce Type: cross Abstract: Applied researchers increasingly convert images into features with pretrained encoders, then use those features in a downstream prediction model. The

Scoring Higher, Answering Worse: Mitigating Reward Hacking in Rubric-Based RL via Protocol-Level Rubrics

arXiv:2609.38847v1 Announce Type: cross Abstract: Rubric-based reinforcement learning (Rubric-RL) trains language models where no verifier exists. A judge checks each criterion of a rubric, and the ve

Structure vs. Chain-of-Thought: Evaluating LLM Criteria Extraction for Depression Severity

arXiv:2609.39049v1 Announce Type: cross Abstract: A large language model (LLM) can rate depression severity directly from a social media post or mark which clinical criteria the post shows and let cod

Synthetic Pre-pretraining Survives Scale, but Not as a Grammatical Prior

arXiv:2609.39827v1 Announce Type: cross Abstract: Pre-pretraining (PPT) on synthetic non-natural language data improves token efficiency during language model pre-training (PT). Prior work attributes

The Geometry of Harmfulness in Multi-Turn Attacks

model-releasesarxiv-cs-cl

arXiv:2609.38389v1 Announce Type: cross Abstract: Large language models (LLMs) remain vulnerable to adversarial attacks that circumvent safety alignment to elicit harmful outputs. It remains unclear h

Local Ai

Understanding Head Geometry and Dynamics in Federated Regression through a Natural Solution Selection Rule: An Unconstrained Feature Model Analysis

arXiv:2609.39464v1 Announce Type: new Abstract: In federated averaging, local objectives can admit multiple optimal heads, making the aggregate depend on which heads clients return. We study this ambi

local-aiarxiv-cs-lg

UniAE-MoE: A Unified Audio Encoder via Mixture of Experts

arXiv:2609.39199v1 Announce Type: cross Abstract: Large Audio Language Models (LALMs) rely on effective audio encoders for multi-task performance. We introduce UniAE-MoE, a unified audio encoder desig

Adapting Linear-Time Architectures for Tabular In-Context Learning

arXiv:2609.36337v1 Announce Type: new Abstract: Tabular foundation models achieve strong performance by conditioning on labelled examples in context, but softmax attention limits their use on large da

researcharxiv-cs-lg

b11279

model-releasesllama-cpp-releases

add GLM-5.3-Flash (GLM5-Next) support (#27773) Rebase GLM-Next support onto master, and migrate to llama-memory-hybrid-idx Add initial MTP support Merge branch optimizations. Reduce allocated compute

cktFormer: Transformer-Based Approach for Automated Analog Circuit Design

arXiv:2609.36752v1 Announce Type: new Abstract: Circuit design is a complex and iterative process that requires expertise in electronic engineering. It involves selecting components while meeting perf

Local Ai

CogWAM: Aligning Semantic Cognition with World Action Modeling via Event-Driven Interfaces

arXiv:2609.37721v1 Announce Type: cross Abstract: Robot policies increasingly incorporate semantic reasoning and future-world prediction, yet combining these capabilities does not guarantee that local

local-aiarxiv-cs-cv

Data Agent Kit is now GA: Bring Google Data Cloud to any coding agent

model-releasesgoogle-cloud-ai

Today, Google Cloud Data Agent Kit is generally available. Data Agent Kit is a free set of Model Context Protocol (MCP) tools and agent skills that lets the coding agent you already use work directly