Search: “llm”

PDF parsing is fun because there's an infinite variety of enteprise documents 📑. For each document category, there's a long tail of work to build more precise bounding boxes, confidence scores, and do

arXiv:2608.26530v1 Announce Type: new Abstract: Long-horizon agent runs generate experience that can improve both the current run and future work. Most self-improvement methods process this experience

arXiv:2608.26735v1 Announce Type: new Abstract: Specializing large language models to vertical domains improves domain-specific behavior but often degrades general capabilities such as reasoning, codi

arXiv:2608.26238v1 Announce Type: new Abstract: Native 3D generators now recover impressive mesh geometry from a single image. However, a dense mesh stays soft where a machined object should be sharp,

arXiv:2608.27394v1 Announce Type: new Abstract: Retrieved scientific literature can serve as inspiration for both human and AI scientists. Inspiration can take different forms: prior work may directly

arXiv:2608.27036v1 Announce Type: new Abstract: Large Language Models (LLMs) can be trained to perform chain-of-thoughts reasoning in order to improve the reliability of their responses. In this work,

arXiv:2608.26732v1 Announce Type: cross Abstract: Graph neural networks (GNNs) are typically conceptualized as message-passing neural networks, yet it remains unclear why neighborhood aggregation reli

arXiv:2608.27417v1 Announce Type: new Abstract: Vision-language models (VLMs) can locate an image region referred to by a text prompt and route the corresponding visual evidence to the output, yet the

arXiv:2608.26829v1 Announce Type: cross Abstract: Time series forecasting models operate on raw numerical sequences, lacking the semantic knowledge that domain experts implicitly leverage, such as the

arXiv:2608.26132v1 Announce Type: cross Abstract: Labeled property graphs combine relational structure with heterogeneous textual and categorical properties attached to both nodes and relationships. C

arXiv:2608.26836v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated remarkable proficiency in natural language understanding, yet they struggle with strict multi-step reason

arXiv:2608.26423v1 Announce Type: cross Abstract: This paper proposes a framework for constructing a classifier as a safeguard layer, and for developing a complementary diagnostic that identifies whic

Hey everyone, I am the co-founder of Tontaube. My brother and I just released TontaubeV1, a 2.9B-parameter open-weight TTS model focused on expressive speech, long-form generation, and low-latency loc

arXiv:2608.26148v1 Announce Type: new Abstract: Depression affects millions worldwide, yet diagnosis relies on subjective self-reports that may miss authentic behavior. This paper presents an approach

arXiv:2608.27128v1 Announce Type: new Abstract: Long-context inference is bottlenecked by the memory footprint of the key-value (KV) cache, especially for small models under tight resource budgets. Ex

arXiv:2608.27311v1 Announce Type: new Abstract: Agent harnesses shape how language-model agents use instructions, tools, and runtime components, but adapting these harnesses requires costly verificati

J'utilise souvent MiniMax H3 avec des images de référence ; je voulais essayer une vidéo que j'avais créée avec Wan (la version plus ancienne), mais cette fois en utilisant seulement une invite (un LL

arXiv:2510.15987v3 Announce Type: replace Abstract: How do inference time and latent computations enable large language models (LLMs) to solve multi-step reasoning problems? We introduce AlgoTrace, a

// Automata from agent traces // How much of your agent's behavior comes from the model, and how much from the harness you wrapped around it? New work collapses an entire corpus of agent traces into a

arXiv:2608.25637v1 Announce Type: new Abstract: Reference-based verifiers are important for evaluating reasoning models and providing accurate outcome rewards in reinforcement learning with verifiable

llama : add --n-cpu-ffn option (#26622) common : dedupe --n-cpu-moe / --spec-draft-n-cpu-moe override loops common : add --n-cpu-ffn to CPU-offload dense FFN weights of first N layers common : general

arXiv:2608.25871v1 Announce Type: new Abstract: Forecasting in large-scale e-commerce marketplaces is increasingly required to support planning: merchants need to evaluate sales outcomes under future

arXiv:2608.26019v1 Announce Type: new Abstract: On-policy self-distillation (OPSD) uses a privileged copy of the student model to provide dense supervision without an external teacher. OPSD keeps this

arXiv:2511.11635v2 Announce Type: replace-cross Abstract: In intelligent education, personalized mathematics question generation aims to produce mathematics questions that satisfy educational requirem

arXiv:2510.16899v2 Announce Type: replace Abstract: AI memory mechanisms primarily focus on preserving information content, often neglecting the validity conditions under which knowledge remains appli

Designing effective reward signals for open-domain question answering is challenging because high-quality responses must simultaneously satisfy multiple aspects of answer quality that are difficult to

arXiv:2601.16905v3 Announce Type: replace Abstract: Machine unlearning in Mixture-of-Experts (MoE) large language models presents a critical yet under-explored challenge. Current unlearning methods ap

arXiv:2608.25934v1 Announce Type: cross Abstract: Automated fact-checking (AFC) systems retrieve evidence and predict claim veracity, yet evaluations omit simple baselines, systems are developed for a

arXiv:2608.25593v1 Announce Type: new Abstract: Agent capability is not determined by the model alone. The agent harness, encompassing memory management, planning strategy, action protocol, and tool/s

arXiv:2608.25449v1 Announce Type: new Abstract: Formal theorem proving enables machine-verifiable evaluation of mathematical reasoning, yet existing benchmarks often emphasize aggregate proof accuracy

Since Qwen's dropped the Qwen4Exp architecture bomb that focus on offloading parameters to n-gram instead of pure mixture of experts, I dug into this and learned quite a lot. Here's the summary. Expec

arXiv:2510.09062v2 Announce Type: replace Abstract: Recent advances in long chain-of-thought (CoT) reasoning have largely prioritized answer accuracy and token efficiency, while overlooking aspects cr

arXiv:2608.25973v1 Announce Type: cross Abstract: Understanding instruction-following capabilities in scientific domains is essential for effectively leveraging Multimodal Large Language Models (MLLMs

arXiv:2608.25990v1 Announce Type: new Abstract: Orthogonal optimisers such as Muon can substantially accelerate large language model pretraining relative to Adam, yet the mechanism remains incompletel

arXiv:2608.26067v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models have demonstrated effectiveness in robot manipulation, yet state-of-the-art models such as pi0.5 operate under a sin

arXiv:2608.25157v1 Announce Type: new Abstract: Automatic AV-sync metrics are widely used to rank and train audio-visual generators, but they are rarely audited as measurement instruments. We jointly

Z.ai Co. today released the code for GLM-5.3-Flash, a large language model that’s 10 times more cost-efficient than its predecessor. The algorithm made its original debut last week under the codename

arXiv:2608.23817v1 Announce Type: new Abstract: SHAP and LIME are now standard tools for interpreting black-box predictions, yet their outputs can vary substantially when the input is perturbed by sma

arXiv:2608.23892v1 Announce Type: cross Abstract: This article presents the abridged core of A Mathematical Theory of Interpretation (MTI), which treats interpretation as observer-relative spectral me

arXiv:2608.24046v1 Announce Type: new Abstract: When an AI algorithm makes decisions that affect more than one person, aligning it becomes a problem of social choice: how should people's divergent pre

arXiv:2601.00282v2 Announce Type: replace-cross Abstract: Quantization is widely used to accelerate inference and streamline the deployment of large language models (LLMs), yet its effects on self-exp

arXiv:2608.24033v1 Announce Type: cross Abstract: Time series classification underpins applications in healthcare, sensing, and industrial monitoring. Although time series foundation models support fo

arXiv:2608.24105v1 Announce Type: new Abstract: Purpose: Automatic radiology report generation (RRG) has been widely explored to improve reporting accuracy and reduce radiologists' workload. Most exis

The Qwen3.8‑Flash‑Next 176 B multimodal MoE model uses a hybrid Gated DeltaNet / Qwen Sparse Attention (QSA) architecture that compresses long‑context history and aggregates tokens into micro‑blocks f

We're releasing a fully quantized NVFP4 version of Qwen3.8-27B. The checkpoint was trained using quantization-aware distillation (QAD) with QUASAR, our new QAT algorithm. We used the original BF16 mod

arXiv:2603.13558v2 Announce Type: replace-cross Abstract: We introduce Holographic Invariant Storage (HIS), a protocol that assembles known properties of bipolar Vector Symbolic Architectures into a d