Knowledge catalogue
Lighthouse Attention is a novel attention mechanism that improves efficiency in transformer models by selectively focusing computation on the most relevant tokens, similar to how a lighthouse beam ill
arXiv:2605.14098v1 Announce Type: cross Abstract: Chain-of-thought (CoT) reasoning with self-consistency improves performance by aggregating multiple sampled reasoning paths. In this setting, correctn
arXiv:2603.23129v2 Announce Type: replace Abstract: Godel agent realize recursive self-improvement: an agent inspects its own policy and traces and then modifies that policy in a tested loop. We intro
arXiv:2605.14294v1 Announce Type: new Abstract: Formal verification of transformers has become increasingly important due to their widespread deployment in safety-critical applications. Compared to cl
arXiv:2603.04885v2 Announce Type: replace Abstract: Real-world dialogue usually unfolds as an infinite stream. It thus requires bounded-state memory mechanisms to operate within an infinite horizon. H
arXiv:2605.14113v1 Announce Type: cross Abstract: While interpretable prototype networks offer compelling case-based reasoning for clinical diagnostics, their raw continuous outputs lack the semantic
arXiv:2605.14502v1 Announce Type: cross Abstract: Power electronics systems are increasingly exposed to cyber threats due to their integration with digital controllers and communication networks. Howe
arXiv:2510.20206v2 Announce Type: replace Abstract: Prompt design plays a crucial role in text-to-video (T2V) generation, yet user-provided prompts are often short, unstructured, and misaligned with t
arXiv:2605.13852v1 Announce Type: cross Abstract: We often aim to generate images that are both photorealistic and 3D-consistent, adhering to precise geometry, material, and viewpoint controls. Typica
arXiv:2605.15178v1 Announce Type: new Abstract: We introduce SANA-WM, an efficient 2.6B-parameter open-source world model natively trained for one-minute generation, synthesizing high-fidelity, 720p,
arXiv:2605.14205v1 Announce Type: new Abstract: LLM-based web agents can navigate live storefronts, yet they often collapse to a single 'average buyer' policy, failing to capture the heterogeneous and
arXiv:2605.14089v1 Announce Type: new Abstract: In recent years, a variety of powerful LLM-based agentic systems have been applied to automate complex tasks through task orchestration. However, existi
arXiv:2605.14284v1 Announce Type: new Abstract: Comparative evaluation of multiple dynamic treatment policies is essential for healthcare and policy decisions, yet conventional longitudinal causal inf
arXiv:2601.23072v2 Announce Type: replace Abstract: Flow matching is a scalable generative framework for characterizing continuous normalizing flows with wide-range applications. However, current stat
arXiv:2605.14708v1 Announce Type: new Abstract: Style-conditioned scene text generation faces unique challenges in extracting precise text styles from complex backgrounds and maintaining fine-grained
arXiv:2601.16981v2 Announce Type: replace Abstract: We present SyncLight, a method to enable consistent, parametric control over light sources across multiple uncalibrated views of a static scene cond
arXiv:2605.15118v1 Announce Type: cross Abstract: We introduce a reusable framework for auditing whether LLM attack benchmarks collectively cover the threat surface: a 4imes6 Target imes Technique mat
arXiv:2605.14136v1 Announce Type: new Abstract: Recent text-to-video diffusion transformers generate visually compelling frames, yet still struggle with temporal coherence, often producing flickering,
arXiv:2605.14477v1 Announce Type: new Abstract: We introduce EvoLib, a test-time learning framework that enables large language models to accumulate, reuse, and evolve knowledge across problem instanc
arXiv:2605.14291v1 Announce Type: cross Abstract: The rapid advancement of Large Vision-Language Models (LVLMs) is increasingly accompanied by unauthorized scraping and training on multimodal web data
arXiv:2605.14705v1 Announce Type: new Abstract: Sign language is the primary language for many Deaf and Hard-of-Hearing (DHH) signers, yet most conversational AI systems still mediate interaction thro
arXiv:2605.14253v1 Announce Type: new Abstract: Purpose: Mechanical thrombectomy (MT) improves stroke outcomes, but is limited by a lack of local treatment access. Widespread distribution of reinforce
arXiv:2603.11042v2 Announce Type: replace-cross Abstract: Generating music that temporally aligns with video events is challenging for existing text-to-music models, which lack fine-grained temporal c
arXiv:2603.18943v2 Announce Type: replace Abstract: This paper presents VGGT-360, a novel training-free framework for zero-shot, geometry-consistent panoramic depth estimation. Unlike prior view-indep
arXiv:2605.15186v1 Announce Type: cross Abstract: High-quality 3D scene reconstruction has recently advanced toward generalizable feed-forward architectures, enabling the generation of complex environ
arXiv:2602.02994v2 Announce Type: replace Abstract: Reinforcement learning has emerged as a principled post-training paradigm for Temporal Video Grounding (TVG) due to its on-policy optimization, yet
arXiv:2605.14607v1 Announce Type: new Abstract: Any new medium, once it emerges, is used for more than the transmission of overt content alone. The information it carries typically operates on two lev
arXiv:2605.13923v1 Announce Type: cross Abstract: We study certified runtime monitoring of past-time signal temporal logic (ptSTL) from visual observations under partial observability. The monitor mus
What we find most useful about Lighthouse is the architectural decoupling. The selection logic is fully separated from the attention kernel, which means every upstream software or hardware attention i
arXiv:2605.14449v1 Announce Type: cross Abstract: Hallucination detection in large language models (LLMs) requires balancing accu racy, efficiency, and robustness to distribution shift. Black-box cons
arXiv:2603.09921v3 Announce Type: replace Abstract: Open-domain visual entity recognition (VER) seeks to associate images with entities in encyclopedic knowledge bases such as Wikipedia. Recent genera
arXiv:2605.14166v1 Announce Type: new Abstract: Face image super-resolution aims to recover high-resolution facial images from severely degraded inputs. Under extreme upscaling factors, fine facial de
arXiv:2605.12586v1 Announce Type: cross Abstract: Vision-language models (VLMs) exhibit a striking paradox: they can generate executable code that reconstructs a 3D scene from geometric primitives wit
arXiv:2605.13142v1 Announce Type: new Abstract: In professional sports, a team has clinched the playoffs if they are guaranteed a postseason spot, regardless of the outcomes of any remaining games. As
arXiv:2605.12622v1 Announce Type: cross Abstract: We formalize action emergence as a target capability for end-to-end autonomous driving: the ability to generate physically feasible, semantically appr
arXiv:2605.12569v1 Announce Type: cross Abstract: Global navigation satellite system (GNSS) interference poses a serious threat to reliable positioning, especially in indoor and multipath-rich environ
arXiv:2605.12954v1 Announce Type: cross Abstract: Long video understanding is heavily bottlenecked by a rigid one-shot paradigm: existing methods either densely encode videos at prohibitive memory and
arXiv:2502.15761v3 Announce Type: replace-cross Abstract: The deployment of large language models (LLMs) on extended reality (XR) devices has great potential to advance the field of human-AI interacti
The search results show recent llama.cpp releases but don't contain specific details about the b9145 build. Based on the pattern of changes in recent releases, b9145 likely includes updates to SYCL/Le
arXiv:2605.13455v1 Announce Type: new Abstract: Synapses are densely packed submicron structures that dynamically reorganize during learning and memory formation. Longitudinal extit{in vivo} imaging o
arXiv:2506.13456v2 Announce Type: replace Abstract: Diffusion Policy has demonstrated strong visuomotor modeling capabilities, but its high computational cost renders it impractical for real-time robo
arXiv:2605.12517v1 Announce Type: cross Abstract: Vision-language models (VLMs) are often deployed on text-only inputs, although they are trained with images. We find that removing the vision modality
arXiv:2605.13295v1 Announce Type: cross Abstract: LLM-based multi-agent systems have demonstrated strong performance across complex real-world tasks, such as software engineering, predictive modeling,
arXiv:2605.12857v1 Announce Type: cross Abstract: Existing API-based agentic systems for RTL code generation are fundamentally misaligned with industrial practice: they assume a golden testbench is av
arXiv:2605.13178v1 Announce Type: cross Abstract: In large vision-language models, visual tokens typically constitute the majority of input tokens, leading to substantial computational overhead. To ad
arXiv:2605.13162v1 Announce Type: new Abstract: Parameter-Efficient Fine-Tuning (PEFT), particularly Low-Rank Adaptation (LoRA), has become a standard approach for adapting Large Language Models (LLMs
Continual Harness proposes an approach to online adaptation for foundation agents that moves beyond traditional gradient-based retraining by introducing a dual-agent architecture (Teacher/Student) wit
arXiv:2605.12809v1 Announce Type: cross Abstract: A critical step for reliable large language models (LLMs) use in healthcare is to attribute predictions to their training data, akin to a medical case
arXiv:2605.13276v1 Announce Type: new Abstract: The rapid evolution of Embodied AI has enabled Vision-Language-Action (VLA) models to excel in multimodal perception and task execution. However, applyi
Deepseek V4 Flash is one of the best 'worker' models out there IMHO so this is very cool! Yet ANOTHER free model available through the Nous Portal 😍😍😍 Can't believe people are still playing with toy l
arXiv:2602.06021v2 Announce Type: replace-cross Abstract: We study a data-dependent notion of diffusion-model generalization: when a model does not memorize the training set, where do its generated sa
arXiv:2605.12939v1 Announce Type: new Abstract: Recent diffusion- and flow-based VTON methods achieve strong results with pretrained generative models, but their reliance on multi-step sampling incurs
arXiv:2605.13296v1 Announce Type: new Abstract: Multi-Agent Path Finding (MAPF) is a coordination problem that requires computing globally consistent, collision-free trajectories from individual start
arXiv:2605.13179v1 Announce Type: new Abstract: The Engram module -- a hash-keyed, O(1) associative memory injected into Transformer layers -- was recently shown to improve large language model pretra
arXiv:2602.03730v2 Announce Type: replace-cross Abstract: Generative foundation models trained on tokenized electronic health record (EHR) timelines show promise for clinical outcome prediction via Mo
arXiv:2605.13462v1 Announce Type: new Abstract: Gesture recognition is a cornerstone of Human-Computer Interaction (HCI) for smart eyewear, enabling natural and device-free control in augmented realit
arXiv:2605.13041v1 Announce Type: new Abstract: With recent advances in embodied agents and AR devices, egocentric observations are readily available as input for real-world interactive online applica
arXiv:2602.05242v1 Announce Type: cross Abstract: Agentic Test-Time Scaling (TTS) has delivered state-of-the-art (SOTA) performance on complex software engineering tasks such as code generation and bu
arXiv:2605.12995v1 Announce Type: new Abstract: Traditional retrieval pipelines optimize utility through stages of candidate retrieval and reranking, where ranking operates over a predefined candidate
Fireworks Training Platform continues to expand. Today GLM 5.1 LoRA RL is now live via Training API: SFT, DPO, and full RL on a 200K context window → custom loss functions or smart defaults. No usage