Search: “inference”

HaloBlock: Faster and Better Decoding for Diffusion Large Language Models

arXiv:2610.17948v1 Announce Type: new Abstract: Diffusion large language models (dLLMs) have attracted increasing research attention due to their bidirectional attention and arbitrary order decoding.

researcharxiv-cs-lg

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

Hardware-Friendly Token-Group Activation Quantization for Low-Bit Mamba Super-Resolution

DGX agent

arXiv:2610.16261v1 Announce Type: cross Abstract: Mamba-based super-resolution models are efficient alternatives to Transformer-based restoration backbones, but their low-bit post-training quantizatio

researcharxiv-cs-cv

14 Oct 2026

inLoRA-MINT: Internal Representations for Auditing Training Data in Domain-adapted LLMs

arXiv:2610.18051v1 Announce Type: new Abstract: This work investigates the application of the Membership Inference Test (MINT) to large language models (LLMs) fine-tuned with LoRA, aiming to determine

Jevolution: A Framework for Studying Communication and Collective Dynamics in AI Agents

arXiv:2610.14449v1 Announce Type: cross Abstract: Experiments with remote decision-making agents depend on more than the policy: the world may change during inference, the simulator determines which a

local-aiarxiv-cs-lg

LongForcing: Scaling Temporal Manifold Matching for Long-Horizon World Rollout

arXiv:2610.18289v1 Announce Type: new Abstract: Autoregressive video diffusion models have emerged as a foundation for long-horizon world rollout. However, when extended far beyond their training wind

safetyarxiv-cs-cv

LoopQuant: Accurate and Efficient Post-Training Quantization for Looped Transformers

arXiv:2610.16368v1 Announce Type: new Abstract: Looped Language Models (LoopLMs) improve parameter efficiency by recursively reusing shared Transformer modules, yet their repeated computation over evo

LoopREG: Representation Supervision Enables Test-Time Recurrent Depth in Diffusion Transformers

arXiv:2610.17877v1 Announce Type: new Abstract: Reusing Transformer blocks offers a way to increase diffusion-model computation at test time, but naive recurrence can degrade generation beyond its tra

local-aiarxiv-cs-cv

Make {pi}0.5 Learn to Call Astra for Reliable Robot Manipulation

arXiv:2610.14117v1 Announce Type: new Abstract: GPT-6 Astra provides robust embodied reasoning, but frequent external queries increase inference time and token costs. We equip a frozen pi_{0.5} policy

safetyarxiv-cs-ro

MFlowHOI: Whole-Body Human-Object Interaction Generation via Part-Aware MeanFlow

arXiv:2610.16252v1 Announce Type: new Abstract: Whole-body human-object interaction (HOI) generation requires coordinating body and finger motion, object trajectories, and time-varying contacts. Exist

local-aiarxiv-cs-cv

MLAfication: Efficiently Enabling High-Fidelity MLA for Modern LLMs

arXiv:2610.17202v1 Announce Type: new Abstract: Multi-Head Latent Attention (MLA) compresses the KV cache into a latent representation, reducing inference memory and bandwidth. Existing MHA2MLA/TransM

Model-Based Online Decision Making via Generative Trajectory Planning

arXiv:2610.18677v1 Announce Type: new Abstract: Online decision-making faces a tension between efficient inference and long-horizon reasoning. Value-based methods rely on short-horizon bootstrapping t

Moving a running model's layers between GPUs mid-reply (8 ms, no re-prefill)

model-releasesr-localllama

I've been working on an inference engine called jitllm. The idea came from my background in operating systems and JIT compilers: generate kernels for whatever hardware is there, and treat layers a bit

Neural Bayesian Networks for Modular and Scalable Decision-Making Under Uncertainty

arXiv:2610.16902v1 Announce Type: cross Abstract: Bayesian Networks (BNs) are established tools for letting agents reason under uncertainty in sequential decision-making. They split a joint distributi

Neuro-Symbolic Graph Transformers for Hierarchical Intention Recognition from Multimodal Human Activity Sequences

arXiv:2610.18176v1 Announce Type: cross Abstract: Proactive human-robot collaboration requires autonomous systems to infer not only immediate actions but also the underlying hierarchical intentions go

agentsarxiv-cs-lg

NeuroVLA: A Brain-inspired Embodied Intelligence for Fluid and Fast Reflexive Robotics Control

arXiv:2601.14628v2 Announce Type: replace Abstract: Vision-language-action (VLA) models inherit powerful semantic representations from large language models, but carrying over their design priorities

local-aiarxiv-cs-ro

Applications

NeuSE: Neural SE(3)-Equivariant Embedding for Long-Term Object-Based Simultaneous Localization and Mapping

arXiv:2303.07308v3 Announce Type: replace-cross Abstract: We present NeuSE, a novel Neural SE(3)-Equivariant Embedding for objects, and illustrate how it supports object-based Simultaneous Localizatio

applicationsarxiv-cs-cv

On the Statistical Mechanics of Attention

arXiv:2610.17347v1 Announce Type: new Abstract: We revisit the statistical mechanics of Transformer attention via modern Hopfield models. We present a free-energy formalism to characterize parameter r

PaceReasoner: Label-Free Learning to Answer Briefly from Long Reasoning

arXiv:2610.17900v1 Announce Type: cross Abstract: Extended chain-of-thought reasoning improves problem-solving performance but increases inference cost and latency, motivating more accurate answers wi

PhySimCode: A Benchmark and Evaluation Method for Physics Video to Code Generation

model-releasesarxiv-cs-cv

arXiv:2610.15722v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) are good at video understanding and code generation, but it remains unclear whether they can recover the physic

Recurrent In-Context Learning for Tabular Foundation Models

arXiv:2610.14998v1 Announce Type: new Abstract: Tabular foundation models adapt to new datasets through in-context learning (ICL), and, like most transformers, they tie computational depth to the numb

Reinforcement Learning In Partially Verifiable Environments for Colon Cancer Treatment Planning

arXiv:2610.17426v1 Announce Type: new Abstract: Clinical decision-making is a sequential process of information collection, clinical inference, and forecasting under partial observability, often witho

Tutorials

Sampling from multimodal distributions with warm starts: Non-asymptotic bounds for the Reweighted Annealed Leap-Point Sampler

arXiv:2512.17977v2 Announce Type: replace-cross Abstract: Sampling from multimodal distributions is a central challenge in Bayesian inference and machine learning. In light of hardness results for sam

tutorialsarxiv-cs-lg

SARATHI: Decoupled Data-Free Probing and Adaptive Reconstruction for Structured LLM Pruning

arXiv:2610.15514v1 Announce Type: new Abstract: Structured pruning reduces the inference cost of large language models (LLMs) by permanently removing neurons and weight rows, yielding dense, hardware-

SEEK: Agentic Knowledge Graph Retrieval for Secure Code Generation

arXiv:2610.14968v1 Announce Type: cross Abstract: Large Language Models (LLMs) can generate functionally correct code while introducing exploitable security weaknesses, particularly when developers sp

agentsarxiv-cs-lg

SP^2ec: Adaptive Self-Speculative Decoding for Vision-Language Models

arXiv:2610.18735v1 Announce Type: new Abstract: Speculative Decoding (SD) has demonstrated its efficacy in accelerating large language model inference, but for Vision-Language Models (VLMs), the diver

Sparse-TD-JEPA: Improving Zero-Shot Reinforcement Learning With Sparse Latent Task Representations

arXiv:2610.14400v1 Announce Type: new Abstract: Successor-feature-based zero-shot reinforcement learning methods pre-train a task encoder and successor-feature predictor on offline, reward-free data,

Teaching MLP More Graph Information: A Three-stage Multitask Knowledge Distillation Framework

arXiv:2403.01079v2 Announce Type: replace Abstract: We study the challenging problem for inference tasks on large-scale graph datasets of Graph Neural Networks: huge time and memory consumption, and t

researcharxiv-cs-lg

TIDE: Trajectory-Informed Skill Distillation for Efficient Text-to-Video Prompt Optimization

arXiv:2610.15098v1 Announce Type: new Abstract: Text-to-video (T2V) models are highly sensitive to prompt specificity, while iterative prompt optimization typically improves generation quality only th

researcharxiv-cs-cv

Applications

TokenPB: Enabling VLA Action Self-Verification via VLM Token Probability Bucketing

arXiv:2603.18091v2 Announce Type: replace Abstract: Test-time sampling can improve Vision-Language-Action (VLA) policies, but only if the model can efficiently select a good action from multiple candi

applicationsarxiv-cs-cv

Toward a Multimodal Foundation Model for Plasma State Reconstruction

arXiv:2610.15076v1 Announce Type: new Abstract: Plasma experiments observe coupled physical states through sparse, heterogeneous, and incomplete diagnostics. We present a proof of concept toward a mul

UniLAP: A Unified Latent Action Policy for Cross-Embodiment Manipulation

model-releasesarxiv-cs-cv

arXiv:2610.16969v1 Announce Type: cross Abstract: Latent Action Models (LAMs) can capture manipulation priors from action-free videos, but discrete latents usually impose a granularity bottleneck, lim

Verify, Act, Distill: Turning Test-Time Compute into Test-Time Adaptation for Vision-Language-Action Models

arXiv:2610.14885v1 Announce Type: new Abstract: Test-time compute (TTC), which scores multiple candidate actions at inference time, is a powerful way to improve the execution performance of vision-lan

safetyarxiv-cs-ro

Quartermaster 1.7 update now lets you run llama.cpp's new decision models (/v1/systemone) with a playground tab

model-releasesr-localllama

Quartermaster (my local inference manager: one server that swaps llama.cpp, image, audio and video models in and out of VRAM) can now help you server and manage decision models: Auto-discovered. Drop

13 Oct 2026

SonicBoom: 52 tok/s decode on a laptop RTX 3050 (llama.cpp: 45) with the same quantized kernels, the difference is a graph compiler. No Python.

model-releasesr-localllama

I’ve been building SonicBoom, a native tensor runtime written in C++/CUDA with a C ABI and no Python anywhere in the stack. I just tagged v0.0.1, and the first target is local LLM inference on consume

13 Oct 2026

A Theory of the Mechanics of Information: Generalization Through Measurement of Uncertainty (Learning is Measuring)

arXiv:2510.22809v2 Announce Type: replace-cross Abstract: Traditional machine learning relies on explicit models and domain assumptions, often limiting flexibility and interpretability. We introduce a

safetyarxiv-cs-ai

AudioTETRIS: Token Elimination via Transitional Importance Signals for Efficient Audio LLMs

model-releasesarxiv-cs-ai

arXiv:2610.13600v1 Announce Type: cross Abstract: Large Audio Language Models (LALMs) achieve strong performance on speech-text tasks, but standard audio encoders produce hundreds of tokens for even s

b11567

model-releasesllama-cpp-releases

llama : report compute graph reallocations during inference (#30329) Add an optional ggml_backend_sched_realloc_callback that user code registers with ggml_backend_sched_set_realloc_callback to be not

CacheScope: Exact-Compute Allocation for Training-Free Diffusion Acceleration

arXiv:2610.12530v1 Announce Type: new Abstract: Training-free feature reuse accelerates diffusion transformer inference by replacing full denoising computation at selected steps with cached approximat

researcharxiv-cs-lg

Contextual Correction of Parallel Draftsfor Speculative Decoding

arXiv:2610.12570v1 Announce Type: new Abstract: Speculative decoding accelerates large language model inference by using a lightweight drafter to propose multiple tokens that are subsequently verified

researcharxiv-cs-cl

Dell and Intel weigh the costs of enterprise AI scalability

Enterprise AI scalability is running into a hard budget wall for some customers. As agentic workloads multiply, cloud token bills are consuming allocations faster than planned, prompting companies to

agentssiliconangle

GeoGR: Spatial-Semantic Token Compression and Group Routing for Efficient 3D VLMs

arXiv:2610.13035v1 Announce Type: cross Abstract: Multi-view representations enable pretrained image and video vision-language models to understand 3D scenes, but their long visual token sequences mak

local-aiarxiv-cs-ai

Hardware

How Much Computation Does Every Forecast Need? Structured Block Skipping in Two Time-Series Foundation Models

arXiv:2610.13208v1 Announce Type: new Abstract: Time-series foundation models use the same network depth for every input, although forecasting windows differ in structure and difficulty. We test wheth

hardwarearxiv-cs-lg

Introducing DigitalOcean Insights: Built-in Observability for Your Agents, Applications, and Infrastructure

Coding agents have changed how fast teams ship, and more of the software they ship now includes agents. A production agent spans model calls, routing decisions, tools, and the infrastructure underneat

agentsdigitalocean

Hardware

Microsoft’s new AI chip makes it less reliant on Nvidia’s GPUs

Microsoft is announcing its third-generation Maia in-house AI chip today, designed to be the 'primary AI inference system' for Microsoft's products and services. Maia 300 not only improves the cost an

hardwarethe-verge-ai

Not All Dimensions Are Equal: Outlier-Aware Attention Linearization with Grouped Quadratic Kernels

arXiv:2610.13674v1 Announce Type: cross Abstract: Linearizing softmax attention offers a route to efficient Transformer inference, but preserving pretrained attention remains challenging. We identify

researcharxiv-cs-ai

Tutorials

One-Step Text Generation by Seq-Drifting

arXiv:2610.12879v1 Announce Type: new Abstract: Diffusion language models generate tokens in parallel, but still require multiple sampling steps to produce high-quality text. In this paper, we investi

tutorialsarxiv-cs-ai

Opening the Black Box of Classifier-Free Guidance via Information Bottleneck

arXiv:2610.13369v1 Announce Type: new Abstract: Classifier-free guidance (CFG) is a standard technique for improving conditional generation in diffusion models, yet its activation schedule is usually

Overcoming Data Imbalances with Expert Selection Biases

arXiv:2610.13370v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) models have become the standard architecture for frontier LLMs. The multiple parallel experts in each transformer block theoret

researcharxiv-cs-cl