Search: “inference”

arXiv:2605.24202v1 Announce Type: new Abstract: Multi-agent LLM workflows route inference through specialized roles to lift end-task accuracy, but jointly training those roles with reinforcement learn

arXiv:2605.23258v1 Announce Type: new Abstract: KV cache growth is a major bottleneck for long-context inference in large language models. Existing methods are often dominated by binary eviction or re

arXiv:2605.23200v1 Announce Type: cross Abstract: The linear growth of the Key-Value (KV) cache is a critical bottleneck in long-form LLM inference. Existing KV compression methods mitigate this by ev

As enterprises confront spiraling cloud inference bills, on-prem AI computing is emerging as the decisive lever for making agentic AI economically viable at scale. Dell Technologies Inc.’s announcemen

arXiv:2510.09450v2 Announce Type: replace Abstract: Low-light video enhancement (LLVE) is challenging due to noise, low contrast, and color degradation. While learning-based methods enable fast infere

arXiv:2605.23096v1 Announce Type: cross Abstract: Fully homomorphic encryption (FHE) enables private inference by evaluating neural networks on encrypted data. In this way, we can delegate the computa

This post discusses how OpenAI is burning 17 billion annually despite 20+ billion in revenue because unit economics for AI inference are fundamentally broken, with the company losing more money on que

arXiv:2605.23652v1 Announce Type: new Abstract: On a 300-persona life-simulation benchmark, pcsp achieves compositional zero-shot persona identification up to 17x above chance, Spearman rho approx 0.7

OpenStudio is a hybrid AI router that combines local model inference with cloud-based model access through OpenRouter, a unified API providing access to hundreds of AI models through a single endpoint

arXiv:2605.23771v1 Announce Type: cross Abstract: Virtual photography asks an agent to enter a prepared 3D scene with no preselected camera pose or reference image, infer a suitable shot from scene in

arXiv:2605.22984v1 Announce Type: cross Abstract: Test-Time Training (TTT) is an emerging paradigm that enables models to adapt their parameters during inference, improving performance on tasks such a

arXiv:2605.23872v1 Announce Type: new Abstract: We introduce training-free looped transformers, in which a lightweight inference-time wrapper loops a contiguous mid-stack block of layers of a frozen c

arXiv:2605.22717v1 Announce Type: cross Abstract: Interactive streaming music generation promises the use of generative models for live performance and co-creation that is impossible with offline mode

arXiv:2605.21780v1 Announce Type: new Abstract: Randomized smoothing is a powerful tool for certifying robustness to adversarial perturbations, including poisoning attacks via randomized training and

arXiv:2602.01279v2 Announce Type: replace Abstract: Bayesian Last Layers (BLLs) provide a convenient and computationally efficient way to estimate uncertainty in neural networks. However, they underes

RWKV-7 G1g is here: the world's best pure RNN LLM, and a competitive LLM in general. Try https://huggingface.co/spaces/BlinkDL/RWKV-Gradio-2 for bsz16 7B inference. G1h in June 🙂 p.s. const 15000+tps

arXiv:2605.22082v1 Announce Type: new Abstract: We present CoRMA(Contrastive Robotic Motor Adaptation), a context-based meta-adaptation framework that modifies RMA for force-dominant assembly. CoRMA r

arXiv:2605.21964v1 Announce Type: new Abstract: Computational imaging enables compact infrared systems, but deep-learning pipelines that combine image reconstruction and object detection often introdu

arXiv:2605.22147v1 Announce Type: new Abstract: High-resolution remote sensing images (RSIs) are crucial for Earth observation applications, yet acquiring them is often limited by sensor constraints a

arXiv:2605.21661v1 Announce Type: cross Abstract: Adapting pretrained diffusion models to downstream objectives such as inverse problems often requires expensive test-time guidance or optimization. We

arXiv:2605.22247v1 Announce Type: new Abstract: Idioms pose a fundamental challenge for language models, as their meaning cannot be inferred from surface form alone. Understanding such expressions, th

arXiv:2602.23200v2 Announce Type: replace-cross Abstract: When transformer-based language models are deployed for text generation, most of the inference time is spent in the decoding stage, where outp

arXiv:2511.07885v4 Announce Type: replace-cross Abstract: Large language model (LLM) queries are predominantly processed by frontier models in centralized cloud infrastructure. Demand growth strains t

The AI PC is being fundamentally redefined as agentic workloads push the boundaries of what local compute can deliver — and as runaway cloud token costs force enterprises to rethink where inference ac

arXiv:2605.22456v1 Announce Type: new Abstract: Cloud-hosted LLM driver agents provide useful semantic judgments, but their inference latency exceeds stepwise vehicle-control windows. Learned world mo

arXiv:2605.22192v1 Announce Type: new Abstract: Blind image quality assessment (BIQA) for ultrahighdefinition (UHD) images remains challenging because native-resolution inference is computationally ex

arXiv:2601.23224v2 Announce Type: replace Abstract: Existing multimodal large language models for long-video understanding predominantly rely on uniform sampling and single-turn inference, limiting th

arXiv:2510.14444v3 Announce Type: replace Abstract: Post-training pruning can substantially reduce LLM inference costs, but it often degrades quality unless the remaining weights are adapted. Since gl

arXiv:2605.20837v1 Announce Type: new Abstract: Architectural spatial intelligence, the ability to recognize and infer architectural space, is fundamental to tasks such as robot navigation, embodied i

arXiv:2602.10455v2 Announce Type: replace-cross Abstract: Driven by scaling laws, recommender systems increasingly rely on larger-scale models to capture complex feature interactions and user behavior

arXiv:2605.20758v1 Announce Type: cross Abstract: Inference-time guided sampling steers state-of-the-art diffusion and flow models without fine-tuning by interpreting the generation process as a contr

arXiv:2605.20936v1 Announce Type: cross Abstract: Hybrid attention architectures are becoming an increasingly important paradigm for improving LLM inference efficiency while preserving model quality,

arXiv:2605.21381v1 Announce Type: new Abstract: Recent advances in Image Restoration (IR) have been largely driven by generative methods such as Diffusion Models and Flow Matching, which excel in synt

arXiv:2605.20950v1 Announce Type: new Abstract: Vision-Language Models (VLMs) face a bottleneck of prohibitive computational costs arising from massive visual token sequences during inference. Existin

arXiv:2605.21007v1 Announce Type: new Abstract: Road segmentation is a fundamental perception task for autonomous driving and intelligent robotic systems, requiring both high accuracy and real-time in

Nathan's @cursor_ai team didn't prompt-engineer their way to Composer 2.5. They trained it. The massive RL program runs RL rollouts on Fireworks, alongside production inference. 'Comment 🔥 to see my p

arXiv:2605.21226v1 Announce Type: new Abstract: The key-value (KV) cache dominates memory bandwidth and footprint in long-context autoregressive inference. Recent rotation-preconditioned codecs (Turbo

arXiv:2603.23890v2 Announce Type: replace-cross Abstract: As the modern microservice architecture for cloud applications grows in popularity, cloud services are becoming increasingly complex and more

arXiv:2605.20813v1 Announce Type: new Abstract: Inference in diffusion large language models (dLLMs) is computationally expensive, as full self-attention must be repeatedly executed at each step of th

arXiv:2605.20868v1 Announce Type: new Abstract: KV cache quantization reduces the memory cost of long-context LLM inference, but introduces approximation error that is typically validated only empiric

arXiv:2605.20234v1 Announce Type: new Abstract: Prior-Data Fitted networks (PFNs) have been very successful in tabular contexts, handling prediction tasks in context. However, they are designed for si

arXiv:2605.20916v1 Announce Type: new Abstract: Implicit sentiment analysis is challenging because sentiment toward an aspect is often inferred from events rather than expressed through explicit opini

arXiv:2603.10139v2 Announce Type: replace Abstract: Every formal grammar defines a language and can in principle be used in three ways: to generate strings (production), to recognize them (parsing), o

arXiv:2506.17631v4 Announce Type: replace Abstract: Time series forecasting aims to model temporal dependencies among variables for future state inference, holding significant importance and widesprea

arXiv:2605.21299v1 Announce Type: new Abstract: Humans effortlessly go beyond literal meanings: If you mow the lawn, I will give you fifty dollars, is typically understood as implying that the speaker

arXiv:2605.20641v1 Announce Type: cross Abstract: Inference optimization is a vital technique for deploying LLMs at scale. Compilation is the most widely adopted optimization technique for LLMs. While