Knowledge catalogue
arXiv:2608.18931v1 Announce Type: cross Abstract: Test-time scaling (TTS) improves language model outputs by spending additional inference compute - generating multiple candidates, searching over part
arXiv:2602.12250v2 Announce Type: replace Abstract: Graph neural networks (GNNs) enable powerful unsupervised learning of communities. However, such inference may inadvertently expose sensitive group
NVIDIA’s Cosmos 3 Edge is a 4‑billion‑parameter omni‑model (with a 2‑billion‑parameter Nemotron reasoner) that can run on an NVIDIA Jetson Thor, providing on‑device policy inference for robot manipula
arXiv:2608.15869v1 Announce Type: cross Abstract: Multimodal large language models increasingly use visual chain-of-thought (Visual CoT) to reason about spatial, temporal, and embodied environments. B
arXiv:2608.14709v1 Announce Type: cross Abstract: This demo presents real-time AI-based uplink channel-estimation inference using data collected from a hardware-in-the-loop 5G platform. The data-colle
arXiv:2608.15018v1 Announce Type: new Abstract: Deploying large language models (LLMs) for inference on edge devices is challenging due to severe memory and bandwidth constraints. While speculative de
Following up on my previous post about my budget server setup (Intel N100 + RTX 5060 Ti 16GB), a few of you asked for a deeper dive into my actual inference config and real-world agentic performance.
arXiv:2608.14003v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) achieve strong performance on complex tasks through extended chain-of-thought generation, but incur substantial computatio
arXiv:2608.14021v1 Announce Type: new Abstract: Transformer-based sequential recommenders with causal self-attention often rely heavily on the most recent interaction at inference time, but how this b
arXiv:2608.13226v1 Announce Type: cross Abstract: While 3D Vision-Language Models (3D VLMs) have demonstrated remarkable spatial reasoning capabilities, they suffer from massive visual token counts th
arXiv:2608.12652v1 Announce Type: new Abstract: Benchmark contamination is diagnosed today with n-gram overlap, with likelihood-based membership inference, or with canary strings, and each needs somet
arXiv:2608.13043v1 Announce Type: new Abstract: Diffusion models have achieved dominant performance in visual generation but suffer from substantial inference overhead. While cache-based acceleration
arXiv:2608.12537v1 Announce Type: new Abstract: Existing 3D facial-landmark methods localize points on visible skin, but whether CT-defined internal skeletal landmarks can be inferred from external so
10,000 @nvidia B300 GPUs. India's largest AI Factory. Together AI and @larsentoubro are building the country's biggest GPU cluster, backing open-source inference, fine-tuning, and training at scale fo
arXiv:2608.11473v1 Announce Type: new Abstract: Inferring the ground-truth from noisy crowdsourced labels is an important theoretical and practical problem. Neural network-based methods offer an alter
arXiv:2608.11768v1 Announce Type: new Abstract: The adaptive neuro-fuzzy inference system (ANFIS) is an interpretable reasoning framework capable of generating explicit IF-THEN fuzzy rules, making it
arXiv:2608.09888v1 Announce Type: cross Abstract: We introduce BDH-CQ, a reasoning model that combines in-context learning with recurrent latent reasoning. Inputs presented at inference time continuou
arXiv:2608.08744v1 Announce Type: cross Abstract: The carbon footprint of any deployed Large Language Model (LLM) accumulates during inference, where repeated use of the model substantially exceeds th
Everything started with the sudden death of my old ASRock J1900. While looking for the perfect ITX replacement, I stumbled upon the Chinese CW-NAS-ADLN-K motherboard, which looked perfect on paper: In
arXiv:2608.09227v1 Announce Type: new Abstract: Omnimodal language models (OLMs) enable unified audio-visual understanding, but processing long joint token sequences makes inference computationally pr
arXiv:2608.07427v1 Announce Type: new Abstract: LLM inference accounts for over 90% of AI operational energy, scaling directly with input token count---a critical inefficiency for telecom network anal
arXiv:2608.06628v1 Announce Type: new Abstract: Diffusion language models (dLLMs) offer a promising alternative to autoregressive models by accelerating inference through parallel decoding. Recent dLL
Spec-dec has been a thing for a while, in fact, it's wasn't an idea that was born for LLM inference. E.g. Uber's https://github.com/uber/submitqueue applied it to a merge queue. Apple & GDM had been r
Not the biggest or shiniest, but it's mine From gaming machine inference on the original llama models, to a 4x RTX 6000 Pro Max Q + 4x 3090s local AI cluster. Pictures are in reverse chronological ord
When you move a model into production, you want the quality you evaluated to carry through the serving stack. @Kimi_Moonshot benchmarked Kimi K3 across major inference providers, and Together AI ranke
arXiv:2602.06337v2 Announce Type: replace-cross Abstract: Causal inference is essential for decision-making but remains challenging for non-experts. While large language models (LLMs) show promise in
Nearly every major AI lab uses Google Cloud infrastructure, including for training of models, inference for agents, and new frontier research. Google Cloud also continues to be the platform of choice
arXiv:2608.04385v1 Announce Type: new Abstract: Vision-Language Models (VLMs) often lose visual grounding during multi-step reasoning: as reasoning chains grow longer, later inference steps rely incre
arXiv:2608.04570v1 Announce Type: new Abstract: Personalized LLMs with persistent memory are increasingly deployed, yet the faithfulness of their user models remains unexamined. We study over-inferenc
arXiv:2608.02791v1 Announce Type: new Abstract: MLLM-based segmentation faces a core segmentation trilemma: high segmentation performance, preserved dialogue ability, and fast inference. Embedding-pre
arXiv:2608.03647v1 Announce Type: cross Abstract: Out-of-distribution (OOD) recommendation is vulnerable to preference shifts induced by a latent environment. Existing methods can infer latent states
arXiv:2608.01651v1 Announce Type: cross Abstract: Hybrid-attention large language models combine full attention with recurrent linear attention to reduce long-context inference costs, yet their autore
arXiv:2608.02123v1 Announce Type: new Abstract: Speculative decoding accelerates LLM inference only when drafted continuations survive target-model verification. Semi-autoregressive drafters such as D
arXiv:2608.00144v1 Announce Type: cross Abstract: Membership inference (MIA) on language models is usually summarised by an aggregate ROC-AUC, but such evaluations are confounded: model-free blind bas
arXiv:2607.29596v1 Announce Type: cross Abstract: Vision-language-action models (VLAs), which leverage the cognition of multimodal information to infer physical-world actions, provide a generalized so
arXiv:2607.27735v1 Announce Type: new Abstract: Speculative decoding alleviates the memory-bandwidth bottleneck in large language model inference, but its acceleration is jointly constrained by drafti
arXiv:2607.27269v1 Announce Type: new Abstract: Multi-head latent attention (MLA) is increasingly important for long-context LLM inference because compact latent states replace the growing key-value (
arXiv:2607.27320v1 Announce Type: cross Abstract: Field-level inference of cosmological initial conditions from galaxy surveys requires a forward model that is simultaneously accurate in the non-linea
Its a custom Swift/Metal inference engine that runs Gemma 4 26B-A4B-IT on M-series Macs with very low RAM. It uses ~2GB instead of ~14 GB. The result is reportedly 5–6 tok/s on an 8 GB M2 MacBook Air
arXiv:2607.24013v1 Announce Type: new Abstract: Production-ready audio-driven avatar generation requires efficient inference without sacrificing fidelity or motion expressiveness. However, existing ac
arXiv:2607.22720v1 Announce Type: cross Abstract: Existing adaptive inference methods for Large Language Models rely on observational heuristics, such as hidden-state similarity or activation magnitud
arXiv:2607.22577v1 Announce Type: new Abstract: Scaling large language models (LLMs) has driven their success, yet dense Transformers couple capacity and computation: every parameter is activated for
arXiv:2607.23488v1 Announce Type: cross Abstract: Text-to-image diffusion models expose many inference-time sampling parameters, including prompts, negative prompts, classifier-free guidance scales, a
arXiv:2607.21843v1 Announce Type: cross Abstract: Empirical Bayes (EB) performs simultaneous inference across many related latent variables. Classical EB assumes that the likelihood p(x | z) is tracta
Etched Inc., a startup with a chip optimized for artificial intelligence inference, today announced that it has raised 300 million in funding. The Series C round was led by Sequoia. SK Hynix Inc., the
arXiv:2607.21042v1 Announce Type: new Abstract: Autoregressive text-to-speech models achieve strong naturalness but suffer from slow inference due to sequential token generation, limiting their deploy
arXiv:2607.21557v1 Announce Type: new Abstract: Modern AI agents rely on elaborate inference harnesses such as Claude Code, Codex, and OpenClaw to drive multi-turn reasoning, tool use, and access to e
I've been building a C99 inference engine from scratch (no Python, no BLAS, just gcc and make) that runs BitNet's ternary models on CPU. A few weeks ago I got obsessed with the matmul kernel - wrote a
arXiv:2607.19082v1 Announce Type: cross Abstract: Autoregressive context models are foundational for learned image compression,but they suffer from slow serial inference. Existing acceleration methods
arXiv:2607.13048v1 Announce Type: cross Abstract: Streaming inference pipelines increasingly pair lightweight fast models with Large Language Models (LLMs) that provide rich semantic understanding at
arXiv:2607.12422v1 Announce Type: new Abstract: Speculative decoding accelerates autoregressive language model inference by using a cheap drafter to propose multiple future tokens and a target model t
arXiv:2607.12358v1 Announce Type: new Abstract: Video diffusion models produce high-quality generations but remain slow at inference due to their sequential denoising procedure. Caching-based accelera
arXiv:2512.01113v2 Announce Type: replace-cross Abstract: Algorithmic reasoning -- the ability to perform step-by-step logical inference -- is a synthetic benchmark for evaluating multi-step reasoning
arXiv:2607.12145v1 Announce Type: cross Abstract: True causal relationships are rarely known, and inferring causal graphs from data is hard. A fundamental challenge is how to assess whether a given ca
arXiv:2607.10541v2 Announce Type: replace-cross Abstract: Sequential recommender systems typically infer user preferences through single-pass encoding of interaction histories without iterative refine
arXiv:2607.11933v1 Announce Type: new Abstract: Cross-encoders achieve high reranking accuracy in Retrieval-Augmented Generation (RAG) pipelines but impose quadratic inference costs that limit real-ti
arXiv:2607.08427v1 Announce Type: cross Abstract: Achieving nanosecond-scale inference latency for deep neural networks (DNNs) has become a primary architectural concern for latency-critical applicati
arXiv:2602.04521v2 Announce Type: replace Abstract: Modern deployments require LLMs to enforce safety policies at scale, yet many controls rely on inference-time interventions that add recurring compu
arXiv:2607.05478v1 Announce Type: new Abstract: Loop invariant inference is a fundamental yet challenging problem in program verification. Recent LLM-aided guess-and-check techniques have shown strong
Massive open-source release! LingBot-Video isn't about video quality; it optimizes for physical reasoning, and it runs sparse: only 3B of 30B params are active at inference. MoE is finally showing up