Knowledge catalogue
arXiv:2610.19605v1 Announce Type: cross Abstract: Long-output inference in large language models suffers from severe KV cache memory and bandwidth bottlenecks. While KV cache quantization alleviates t
arXiv:2610.18945v1 Announce Type: cross Abstract: Amortized Bayesian inference uses simulations to perform Bayesian estimation when likelihoods are expensive or analytically unavailable. However, lear
arXiv:2610.14352v1 Announce Type: new Abstract: Signal Temporal Logic (STL) inference provides interpretable formulas for time series classification. For multiclass classification, however, accurate p
arXiv:2610.15207v1 Announce Type: new Abstract: Large Language Models (LLMs) have been shown to be computationally expensive and vulnerable to jailbreak attacks, in which adversarial prompts bypass sa
arXiv:2510.17063v2 Announce Type: replace-cross Abstract: Mean-field variational inference (MFVI) is a widely used method for approximating high-dimensional probability distributions by product measur
arXiv:2610.17579v1 Announce Type: cross Abstract: Inference-time alignment steers diffusion models toward downstream objectives without costly fine-tuning, but existing methods typically optimize expe
arXiv:2610.13844v1 Announce Type: new Abstract: Generative molecular design typically relies on over-generation and post-hoc filtering. Inference-time steering methods like Feynman-Kac (FK) steering i
AI inference is fast becoming the workload that decides the economics of the AI boom. Training built the first wave of GPU clouds, but serving models faster and cheaper will define the next. That shif
arXiv:2610.11923v1 Announce Type: cross Abstract: The brain maintains stable cognition despite continuously changing neural activity. How to extract stable cognitive states from variable neural observ
arXiv:2610.11223v1 Announce Type: cross Abstract: Large Reasoning Language Models (LRLMs) enable multi-step reasoning for robotic task planning, but continued reasoning can overwrite valid intermediat
arXiv:2610.11456v1 Announce Type: new Abstract: Two-photon calcium imaging is a standard tool for recording large neural populations in vivo, yet inferring spikes accurately across the growing diversi
arXiv:2610.09424v1 Announce Type: cross Abstract: Deploying Mixture-of-Experts (MoE) models relies heavily on Expert Parallelism, which generates intense inter-GPU communication. Consequently, state-o
arXiv:2610.09250v1 Announce Type: new Abstract: Best-of-N (BoN) is a common inference-time alignment method that selects the highest-scoring response among N samples from a reference model. Evaluating
arXiv:2610.09032v1 Announce Type: new Abstract: Perceiving structured shapes, such as human faces, from pixels is an inherently ambiguous task in real-world conditions. Yet, shape inference is largely
arXiv:2610.10317v1 Announce Type: new Abstract: Tabular foundation models (TFMs) are pretrained across diverse tabular tasks and make predictions on a new table at inference time using its labeled exa
arXiv:2610.07219v1 Announce Type: new Abstract: Mixture-of-experts models make nearly trillion-parameter capacity accessible with sparse per-token computation, provided that the serving system can dis
arXiv:2610.07389v1 Announce Type: cross Abstract: Sparse autoencoders are widely used to uncover interpretable features in neural networks, yet reliable recovery remains difficult when features overla
arXiv:2610.04681v1 Announce Type: new Abstract: Vision-language-action (VLA) and world-action models (WAMs) often degrade under out-of-distribution task variations despite retaining partial task capab
arXiv:2610.05684v1 Announce Type: cross Abstract: Modern deep generative models are primarily studied for their ability to generate realistic samples, yet the generative dynamics they learn can also s
arXiv:2610.02245v1 Announce Type: cross Abstract: Amortized simulation-based inference (SBI), which is trained on radiative-transfer simulators, recovers exoplanet atmospheres accurately on synthetic
Working on a resource dashboard, rich logs, lightweight 64mb cap, all local, scales on network API endpoints via collector, supports multiple engines (llama, strata, custom cuda engines, unsloth, LMS)
arXiv:2610.01933v1 Announce Type: cross Abstract: Inference-time scaling adapts pretrained diffusion models to new sampling tasks without additional training. Existing methods rely primarily on Monte
arXiv:2610.00907v1 Announce Type: cross Abstract: Relational foundation models (RFMs) are pretrained once on a collection of relational databases and prediction tasks, and then applied zero-shot to pr
arXiv:2508.06980v2 Announce Type: replace Abstract: Recent and rapid advances in artificial intelligence (AI) make it increasingly important to understand the foundations of adaptive behaviour in auto
arXiv:2609.38823v1 Announce Type: new Abstract: Multimodal mixture-of-experts (MoE) models combine sparse expert activation with visual-language capabilities, yet their inference remains costly becaus
arXiv:2609.39164v1 Announce Type: new Abstract: Score-based variational inference (VI) provides an alternative to Kullback--Leibler (KL)-based VI by minimizing the Fisher divergence between the variat
arXiv:2602.08005v2 Announce Type: replace-cross Abstract: Efficient long-context inference faces two coupled bottlenecks: KV-cache memory grows linearly with context length, while attention computatio
arXiv:2609.38414v1 Announce Type: new Abstract: Synthetic data generation is dominated by the fit-then-sample paradigm: a generative model is trained on a private dataset and then sampled from. Despit
arXiv:2609.36307v1 Announce Type: new Abstract: Partial Least Squares (PLS) regression extracts a few outcome-aligned directions in a high-dimensional X and is widely used across applied science, but
arXiv:2609.35882v1 Announce Type: cross Abstract: Genome foundation models are most useful where sequences are generated, yet the largest models need datacenter accelerators and a place to send privat
Amazon Bedrock now supports Anthropic's Claude Opus 5 and Claude Sonnet 5 with in-region inference in Seoul, and Claude Sonnet 5 in Singapore. If you have local data processing requirements in South K
arXiv:2609.36396v1 Announce Type: cross Abstract: As black-box machine learning models become increasingly common, extracting interpretations with uncertainty quantification has become a critical chal
arXiv:2609.36471v1 Announce Type: cross Abstract: World-Action Models (WAMs) improve robotic manipulation by conditioning action generation on predicted future observations, but future prediction adds
arXiv:2609.37902v1 Announce Type: new Abstract: Existing LLM routers choose among models using static per-model costs. We show that open-weight inference markets introduce a second, largely ignored de
arXiv:2609.34818v1 Announce Type: cross Abstract: AlphaFold2 is written in JAX, so the same inference code compiles and runs unchanged on CPUs, GPUs and Google Cloud TPUs. That portability makes the a
arXiv:2609.32204v1 Announce Type: cross Abstract: Secure Transformer inference protects sensitive inputs but incurs substantial cryptographic overhead, with nonlinear operations such as Softmax and Ge
arXiv:2609.32663v1 Announce Type: new Abstract: The memory usage and decoding latency of LLM inference grow rapidly with context length. To reduce these costs, key-value (KV) cache compression methods
arXiv:2601.22132v2 Announce Type: replace Abstract: Large Language Models (LLMs) deliver state-of-the-art performance on complex reasoning tasks, but their inference costs limit deployment at scale. S
arXiv:2609.32444v1 Announce Type: new Abstract: We study training-inference mismatch in reinforcement learning with verifiable rewards (RLVR) for large language models, where rollouts are sampled by a
arXiv:2609.34370v1 Announce Type: new Abstract: Spexis is a multi-GPU LLM inference framework that improves the efficiency of pipeline and tensor parallelism through speculative parallelism. Rather th
arXiv:2609.34130v1 Announce Type: new Abstract: This monograph develops a unified account of training and inference in Power Law Decoder Representation language models (PLDR-LLMs). Exact finite work i
arXiv:2609.30474v1 Announce Type: new Abstract: Speculative decoding is a successful technique speeding up inference of a target autoregressive language model via a fast drafter model. Lossy speculati
arXiv:2609.29466v1 Announce Type: cross Abstract: Approximate message passing on factor graphs underlies two dominant families of probabilistic inference algorithms: expectation propagation (EP) and v
arXiv:2609.29029v1 Announce Type: new Abstract: Onboard vision-language models could enable satellites to answer queries directly, but exhaustive tiled inference over high-resolution imagery is slow a
arXiv:2609.29216v1 Announce Type: cross Abstract: Diffusion models are powerful priors for linear inverse problems, but the reference guidance methods, Diffusion Posterior Sampling (DPS) and Pseudoinv
arXiv:2609.29505v1 Announce Type: new Abstract: Diffusion inference repeatedly evaluates the same large network. We ask whether pretrained weights alone can identify residual branches that need not be
arXiv:2609.28605v1 Announce Type: cross Abstract: The Factivity Inference Evaluation 2026 (FIE2026) classifies Chinese context-hypothesis pairs into nine ordered factivity intervals. Its evaluation me
arXiv:2609.27373v1 Announce Type: new Abstract: Recurrent language models repeatedly apply shared network blocks to refine latent representations, but standard inference recomputes global attention at
arXiv:2609.28064v1 Announce Type: new Abstract: Driving world-action models improve planning by coupling multimodal reasoning with future prediction, but their growing inference cost increasingly conf
arXiv:2609.26300v1 Announce Type: new Abstract: Despite their strong performance, large language models (LLMs) are bottlenecked by KV cache memory traffic during long-context inference. Sparse attenti
Robots are getting smarter, but how can their hardware match that growth? New Microsoft Research findings show that moving AI inference beyond the robot can improve task success, boost efficiency, and
arXiv:2609.26061v1 Announce Type: cross Abstract: Mixture-of-experts (MoE) models improve capacity with moderate overhead by sparsely activating experts per token. However, deploying MoE across resour
arXiv:2609.24422v1 Announce Type: new Abstract: Hierarchical data is ubiquitous in the empirical sciences and is most commonly analyzed with generalized linear mixed-effects models (GLMMs). Bayesian i
arXiv:2602.24208v2 Announce Type: replace Abstract: Diffusion models achieve state-of-the-art video generation quality, but their inference remains expensive due to the large number of sequential deno
arXiv:2602.01654v2 Announce Type: replace Abstract: Steering vectors (SVs) offer a lightweight way to control large language models (LLMs) at inference time by shifting hidden activations, providing a
arXiv:2609.24274v1 Announce Type: new Abstract: Deploying language-conditioned manipulation without a dedicated GPU requires efficient inference and action chunks that cover the delay between policy q
Qwen3.8-27B, up to 144 tok/sec on M5 Max. Let that sink in. ⚡️✨🚀 Excited to partner with @inco_ai to bring you their Splash inference engine in LM Studio on day 0! Get it running now: https://lmstudio
arXiv:2609.19539v1 Announce Type: cross Abstract: Active subspace methods provide a framework for quantifying predictive uncertainty in high-dimensional models by identifying and performing inference
A global fintech uses Together’s Dedicated Model Inference (DMI) to run its GLM‑5.2 coding assistant, handling spike‑heavy traffic that peaks during engineering hours and cannot be met by static capac
arXiv:2609.17837v1 Announce Type: cross Abstract: Hybrid domain decomposition methods provide a flexible framework for coupling full order models (FOMs) and reduced order models (ROMs), but typically