Search: “context”

arXiv:2604.02441v2 Announce Type: replace Abstract: Hybrid state estimators that combine model-based Kalman filtering with learned components have shown promise on simulated data, yet their performanc

Are agent skills always worth using? The answer is no? This paper provides some important insights to understand this more. (bookmark it) Paper summary: Adding procedural skills to an agent is usually

arXiv:2607.22264v1 Announce Type: new Abstract: Autoregressive foundation models trained on tokenized electronic health records (EHRs) can support zero-shot clinical prediction, yet most operate on st

common: fix explicit -md precedence over draft sidecar resolution (#26165) common: fix explicit -md precedence over draft sidecar resolution Follow-up of #25955, an explicit --model-draft file given w

Big update: Among open-weight models, Kimi K3 (Max) is #1 in the Agent Arena with +9.75% net-improvement, surpassing GLM-5.2 (Max) at +7.12%, and landed the #1 spot across 5 signals (see below). Kimi

arXiv:2607.22300v1 Announce Type: new Abstract: We present a systematic study of healthcare-domain cross-lingual transfer to address the scarcity of biomedical NMT resources for Arabic-script language

arXiv:2607.22139v1 Announce Type: new Abstract: Accurate pixel-level classification of coronary angiograms is critical for cardiovascular disease assessment, yet the field lacks standardized evaluatio

arXiv:2607.21848v1 Announce Type: new Abstract: Recent conditional video generation models have shown promising potentials to transform 3D engine renderings, such as depth maps and untextured geometry

arXiv:2607.21861v1 Announce Type: new Abstract: We study baking documents directly into the weights of a 4-bit Gemma-4-e4b model via LoRA, so a system can answer questions about a corpus closed-book:

arXiv:2607.22041v1 Announce Type: new Abstract: There is a growing need for reliable and culturally validated instruments to assess psychological dependency on large language models (LLMs), particular

arXiv:2607.10428v2 Announce Type: replace Abstract: Evaluating large language models (LLMs) as multi-turn conversational partners requires probing capabilities that single-turn benchmarks miss: person

arXiv:2607.22016v1 Announce Type: new Abstract: MEMEs are widely used on the internet and often carry strong elements of sarcasm or irony. Understanding their hidden meanings typically requires a join

arXiv:2607.22302v1 Announce Type: new Abstract: Reconstructing dynamic human faces from brain activity provides a powerful way to study how the mind perceives identity, expression, and facial motion.

Happy to have @FireworksAI_HQ as our day0 launch partner and bring Kimi K3 to more developers. With Fireworks, you can now deploy and fine-tune the 2.8T Kimi K3 model with just a few clicks! Kimi K3 i

arXiv:2607.22361v1 Announce Type: new Abstract: We study information bottlenecks in modern deep-learning architectures -- RNNs, softmax transformers, linear-attention transformers and state-space mode

arXiv:2607.22153v1 Announce Type: cross Abstract: Industrial health management increasingly relies on heterogeneous information sources, including condition monitoring systems, supervisory control and

arXiv:2510.24668v2 Announce Type: replace Abstract: Language agents have demonstrated remarkable potential in web search and information retrieval. However, many search-agent benchmarks assume that us

arXiv:2607.22351v1 Announce Type: new Abstract: The speed of sound in tissue is a prerequisite for well-focused imaging and has diagnostic value, but recovering it from raw pulse-echo channel data is

arXiv:2607.22380v1 Announce Type: new Abstract: Efficient processing is becoming increasingly important in infrared remote sensing, where satellite constellations produce large volumes of observations

Kimi K3 is now available for inference & training in @FireworksAI_HQ. Crazy how easy they make it to tune frontier open models like K3 using LoRA adapters. Best time to figure out how to own your inte

Kimi K3 is now available on @digitalocean 's Serverless Inference! Developers can start building with our most capable model in minutes. .@Kimi_Moonshot K3 from Moonshot AI is now live on DigitalOcean

Kimi K3 is now available on Ollama’s cloud. To use it with Claude Code, run: ollama launch claude --model kimi-k3:cloud Currently Kimi K3 requires a Pro or Max subscription, and consumes extra usage c

arXiv:2607.21856v1 Announce Type: cross Abstract: Modern reasoning models depend on reasoning data, today sourced from human annotations or distilled from stronger LLMs. However, a rich and largely un

arXiv:2607.15456v2 Announce Type: replace-cross Abstract: Looped, weight-tied Transformers reduce parameters by reusing a single block, but decoding still stores a separate K/V cache for every recurre

love this frame. calls to mind the role of the hippocampus in human navigation (via place cells and grid cells), and how navigation is, in a sense, what makes agents *agents* vs plain old LLM calls in

arXiv:2606.18681v2 Announce Type: replace Abstract: Despite their remarkable performance, Vision Language Models (VLMs) incur substantial computational overhead due to the large number of visual token

NVIDIA’s Nemotron 3 Ultra, when paired with the ACE‑RTL agent, achieves a 97.1 % average pass rate on the CVDP benchmark across nine RTL task categories—surpassing GLM 5.2 and Kimi K2.6 while using up

arXiv:2607.21670v1 Announce Type: cross Abstract: Action tokenization maps continuous robot action chunks to discrete tokens and has become an important interface for modern visuomotor policies. Exist

arXiv:2607.21820v1 Announce Type: cross Abstract: Audio deepfake detectors are trained to distinguish genuine speech from synthetic speech and often perform well on standard benchmarks. Yet the same d

arXiv:2607.22413v1 Announce Type: cross Abstract: Sample retrieval tools can help composers find harmonically compatible material, but querying from a fixed reference sample becomes less informative a

arXiv:2607.22535v1 Announce Type: cross Abstract: Action-conditioned video world models predict future observations from an initial observation and an action signal. In robotics, actions influence fut

arXiv:2607.21850v1 Announce Type: new Abstract: As semiconductor manufacturing advances toward sub-2nm nodes, local place-and-route (P&R) design-rule violation (DRV) fixing is increasingly limited by

arXiv:2607.22043v1 Announce Type: new Abstract: Although large language models (LLMs) exhibit remarkable reasoning capabilities, their reliance on text-only pre-training restricts the perception of th

arXiv:2607.22332v1 Announce Type: new Abstract: Interactive deep image segmentation enables efficient medical image annotation by iteratively refining predictions from user prompts, such as positive a

arXiv:2607.22231v1 Announce Type: new Abstract: Video super-resolution (VSR) using large-scale Diffusion Transformer (DiT) priors achieves exceptional perceptual quality but is often impractical due t

Instead of 2T+ models, continuing to release highly capable small to medium size LLMs would really help to keep this community vibrant. Hardly anyone can even dream of running the recent 1.5-2T+ beast

arXiv:2505.19212v2 Announce Type: replace Abstract: Recent advances in LLMs have enabled their use in complex agentic roles, involving decision-making with humans or other agents, making ethical align

When I released Inflect v2 last week, I thought most people would ask whether a TTS model this small actually sounded decent. Instead, I kept getting two questions: “Can I train it on my own voice?” “

Enterprise investment in agentic artificial intelligence is accelerating, but the infrastructure supporting those systems is still catching up. Organizations are moving agents into customer support, s

Last week someone here said ThinkingCap and Fable Fusion 'really do beat the OG' for agentic work, so I ran it: 6 self-grading tasks, 5 reps, 3 models, 90 isolated runs. Tooling, since that's half the

Qwen3.6-27B: SFT vs continued pre-training vs RL? I’m interested in adapting Qwen3.6-27B, but I’m increasingly unsure whether conventional SFT/LoRA is the best route if the goal is to add a capability

I have been running some experiments with smaller open-weight LLMs on multiple-choice questions of Swedish medical licensing exams. On a dataset called MedQA-SWE, GPT-4 scored 84% accuracy in 2024 and

I have tried ollama locally on my gaming PC (5070 with 12bg of VRAM) It works pretty nice on qwen2.5-coder:7b (and 14b) So I decided to take a step further, and install an openWebUI instance on my hom

Yes, open-source / open-weight models are important for a healthy AI ecosystem. That's how we can verify things, check claims, and keep up outside the closed labs. Plus, it gives us the freedom to run

I’ve spent the past month trying to find the point where an extremely small TTS model stops feeling like a size experiment and starts feeling genuinely useful. Today I’m releasing Inflect v2, with two

I ported Krea 2 Identity Edit support to Forge Neo as an extension — the instruction-based, identity-preserving edit LoRA that until now required ComfyUI + custom nodes. It replicates the full dual-co