Knowledge catalogue
Ben Jiang / South China Morning Post: DeepSeek details DSpark, a speculative decoding framework for its V4 models, saying it speeds up AI inference by up to 85% and was tested on Gemma and Qwen — Chin
arXiv:2606.27455v1 Announce Type: cross Abstract: We address the problem of inferring a directed network from nodal measurements generated by linear diffusion dynamics on the sought graph. Observation
Going to be in Paris for @RaiseSummit? Join us and @nvidia to hear from leaders at both companies on the future of AI inference and infrastructure. After that: cocktails and a DJ. Oh la la! See you th
arXiv:2511.20687v3 Announce Type: replace-cross Abstract: This paper presents a novel hybrid approach for coupling subdomain-local non-intrusive Operator Inference (OpInf) reduced order models (ROMs)
arXiv:2507.06722v2 Announce Type: replace Abstract: Understanding how large language models (LLMs) internally represent and process their predictions is central to detecting uncertainty and preventing
This is from a popular inference provider GLM-5.2 plus the US banning the most capable new models means open source caught up to SOTA closed source coding models This could be v problematic for Anthro
This is very interesting. Coinbase seems to have lowered their token spend ($$) to about half, by 1) routing to cheap inference like GLM 5.2 and Kimi 2.7 that are still pretty performant 2) Smart rout
arXiv:2606.27090v1 Announce Type: cross Abstract: Global objectives, such as KL divergence and ELBO, are widely used in Bayesian inference for measuring distributional discrepancy. This paper studies
arXiv:2606.26257v1 Announce Type: new Abstract: How much of my data was used to train a machine learning model? Dataset Usage Inference (DUI) aims to answer this by estimating what fraction of a datas
Fireworks AI is now live on EvoSkill v1.3.0! You can now use @FireworksAI_HQ directly with EvoSkill to run fast inference on open models as both the evolution harness backend and the LLM scorer. Along
arXiv:2606.26588v1 Announce Type: new Abstract: A central challenge in deploying learned robot policies is inference-time behavior steering: redirecting a policy at test time to satisfy user preferenc
arXiv:2505.20178v2 Announce Type: replace-cross Abstract: Prediction-Powered Inference (PPI) is a popular strategy for combining gold-standard and possibly noisy pseudo-labels to perform statistical e
arXiv:2606.26476v1 Announce Type: cross Abstract: Warm-started diffusion samplers accelerate iterative inference, but it is rarely clear which part of the pipeline carries the gain. We study extbf{ret
arXiv:2510.17059v2 Announce Type: replace Abstract: Zero-shot imitation learning requires an agent to reproduce expert behavior from a single demonstration without additional environment interaction o
arXiv:2606.25745v1 Announce Type: cross Abstract: Mean Field Variational Inference (MFVI) is widely understood to underestimate posterior variance. By analysing conjugate Bayesian Linear Regression (B
arXiv:2606.25986v1 Announce Type: new Abstract: We study whether a scaling-law-style inference-compute frontier appears in limit order book prediction. Using FI-2010 and a suite of models ranging from
arXiv:2606.25265v1 Announce Type: new Abstract: Particle-based variational inference (ParVI) methods approximate an intractable target distribution by evolving an ensemble of interacting samples. Exis
OpenAI and Broadcom unveiled Jalapeño, OpenAI's first Intelligence Processor designed specifically for LLM inference , with early testing showing substantially better performance per watt than current
OpenAI: OpenAI and Broadcom unveil Jalapeño, an LLM-optimized inference chip developed from design to manufacturing tape-out in nine months, aided by OpenAI's models — Loading... - Early testing shows
The hard part of reinforcement learning on a frontier model is the infrastructure that keeps training and inference numerically identical: zero KLD, end to end. We've solved this challenge, and are no
New research explores how game theory can optimize NVIDIA Dynamo disaggregated serving architectures, revealing critical performance thresholds and reducing worst-case AI inference latency by up to 7.
.@cartesia runs one of the hardest inference workloads: real-time voice. Their stack has to keep long-lived streams moving, serve millions of audio minutes a day, and hold model latency around 90ms. T
arXiv:2504.08061v2 Announce Type: replace Abstract: Traffic prediction is difficult due to the complex interplay of temporal evolution, spatial interactions, and delayed spatio-temporal propagation ov
Modal Auto Endpoints is a feature that automatically optimizes and manages inference endpoints, allowing users to deploy machine learning models with reduced operational overhead. The service likely h
arXiv:2606.20852v1 Announce Type: new Abstract: Inference-time engineering can alter model behavior without fine-tuning. However, its utility for improving diagnostic performance in medical vision-lan
The sharpest questions in AI live at the local–cloud boundary : where should inference run, & where should your data live? In town for AIE? Come hash it out with @theoryvc, @lancedb & @ollama. June 30
arXiv:2606.12342v1 Announce Type: cross Abstract: Domain fine-tuning degrades the safety of large language models: fine-tuned specialists readily comply with harmful prompts framed in domain language.
arXiv:2508.17077v3 Announce Type: replace-cross Abstract: Current experimental scientists have been increasingly relying on simulation-based inference (SBI) to invert complex non-linear models with in
arXiv:2606.11473v1 Announce Type: cross Abstract: Prior-fitted networks (PFNs) are a promising class of tabular foundation models that perform in-context learning, whereby the entire labelled training
arXiv:2505.00571v3 Announce Type: replace-cross Abstract: Machine Learning (ML) is gaining popularity in epidemiology and healthcare studies for hypothesis-free discovery of risk and protective factor
arXiv:2505.17623v2 Announce Type: replace-cross Abstract: Verifiable computing (VC) has gained prominence in decentralized machine learning systems, where resource-intensive tasks like deep neural net
arXiv:2606.11562v1 Announce Type: cross Abstract: Graph analysis underlies many applications whose answers cannot be looked up in a single record or retrieved along a path: laundering rings, drug repu
arXiv:2605.06485v2 Announce Type: replace-cross Abstract: Large language models (LLMs) have transformed artificial intelligence, but their computational requirements remain prohibitive for most users.
arXiv:2606.11196v1 Announce Type: cross Abstract: Decentralized LLM inference networks need lightweight, reference-free quality evaluation for Proof of Quality (PoQ). We present PoQ-Judge, a framework
arXiv:2606.11201v1 Announce Type: cross Abstract: The wide deployment of LLMs has made model alignment necessary to make newly trained models safely and effectively respond to user instructions. Among
arXiv:2606.10493v1 Announce Type: cross Abstract: Local deployment of large Mixture-of-Experts (MoE) models falls short of the service quality achieved in cloud-scale environments, even under low-conc
arXiv:2606.10217v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly fine-tuned on domain-specific datasets that may contain sensitive, dataset-level properties. Recent work h
arXiv:2606.11186v1 Announce Type: new Abstract: Low-light video enhancement (LLVE) remains a challenging task due to severe information degradation under low-illumination conditions. Recent multimodal
arXiv:2606.10935v1 Announce Type: cross Abstract: Large language model inference is bottlenecked by autoregressive decoding, where each token requires a full forward pass. Multi-token prediction (MTP)
Learn how @cursor_ai partnered with Together AI to deliver real-time inference for AI-powered coding in this article from @ce_zhang and @realDanFu. Cursor's in-editor agents generate code while develo
arXiv:2606.09879v1 Announce Type: new Abstract: This study addresses on-device inference bottlenecks of Transformer models on Tenstorrent's Tensix architecture and proposes an operator fusion strategy
arXiv:2606.09937v1 Announce Type: cross Abstract: We introduce RKSC (Reasoning-Aware KV Cache Sharing), a training-free inference framework that eliminates two structural redundancies in multi-branch
arXiv:2606.08742v1 Announce Type: new Abstract: Abnormality detection is a crucial yet challenging task in medical image analysis. Distinguishing abnormalities from normal data by learning to reconstr
arXiv:2606.09080v1 Announce Type: new Abstract: Pruning has emerged as a dominant paradigm for accelerating large language model (LLM) inference, spanning a broad spectrum of methods that remove compu
arXiv:2606.09514v1 Announce Type: new Abstract: Large language models (LLMs) incur high inference cost due to their depth and parameter scale. Depth pruning can reduce latency by skipping redundant Tr
arXiv:2606.08962v1 Announce Type: cross Abstract: World Action Models (WAMs) generalize better than standard Vision-Language-Action (VLA) policies to novel motions and environments, because a video-mo
arXiv:2510.10028v2 Announce Type: replace-cross Abstract: The rapid advancement of Low-Altitude Economy Networks (LAENets) has enabled a variety of applications, including aerial surveillance, environ
arXiv:2606.09508v1 Announce Type: new Abstract: Existing sparse attention and KV cache compression methods for long-context LLM inference typically apply fixed sparsity patterns or uniform budgets acr
arXiv:2606.08831v1 Announce Type: new Abstract: Large language models (LLMs) increasingly perform multi-step reasoning, where intermediate claims form implicit directed acyclic graphs whose node corre
arXiv:2606.08850v1 Announce Type: cross Abstract: Inference-Time Scaling (ITS) has largely succeeded in verifiable domains like math and coding, where cheap verification enables scalable output select
arXiv:2606.07624v1 Announce Type: new Abstract: This discussion argues that sequential statistical inference can naturally contribute to LLM trustworthiness. In deployment, LLM systems are queried rep
arXiv:2606.08854v1 Announce Type: cross Abstract: Standard Reinforcement Learning with Verifiable Rewards (RLVR) training allocates a fixed rollout budget to every query, without regard for what each
Three MLX videos dropped at WWDC: Running agents locally by @angeloskath https://www.youtube.com/watch?v=wykPErJ8M-8 Distributed inference and training by Tatiana Likhomanenko https://www.youtube.com/
arXiv:2501.06659v2 Announce Type: replace-cross Abstract: Many templatized documents are programmatically generated from structured data following a visual template. Such documents include invoices, t
arXiv:2606.08094v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) policies are typically shipped as Python/PyTorch stacks that assume a workstation-class GPU, a mismatch for the hardware
arXiv:2606.08098v1 Announce Type: new Abstract: Majority voting over sampled answers is the dominant unsupervised aggregator for multi-sample LLM inference. We show that piping the signals every sampl
Introducing 'Signals of AI' - DeepAgent powered by - @nebiustf for Inference - @NVIDIAAI Nemotron 3 Ultra - @tavilyai for Web Search - @LangChain for the core deepagent architecture! An ideal engine f
We are releasing our first quantized checkpoints for the Qwen3.5 series of models, co-designed jointly with our inference engine to achieve maximum possible performance on Apple hardware Starting from
arXiv:2606.05332v1 Announce Type: new Abstract: Patch-based Time Series Foundation Models (TSFMs) suffer from context poisoning: structurally anomalous patches capture disproportionate attention and s
Four weeks, three enterprise customers, same pattern across all three. Most inference spend goes to retrieval loops. A generic index carries no knowledge of domain, query types, or task structure, so