S$^4$R is a method for compressing the Key-Value cache in large language models by building low‑rank subspaces from selectively sampled tokens and performing attention over a sparsely reconstructed KV representation. It initializes key/value bases using a representative prompt subset, reducing reliance on external calibration data while avoiding the high compute cost of full‑prompt decomposition. Experiments on LongBench and RULER with Llama and Qwen models demonstrate up to 5× KV compression with near‑full‑cache accuracy, blending the efficiency of fixed compression with the adaptability of prompt‑dependent approaches.
By Jialong Han, You Wu, Kewei Tu
arXiv:2609.22702v1 Announce Type: new
Abstract: Accurate flood forecasts several days in advance are essential for flood control, water-resource management, and emergency response. Producing them at...
By Hong Zhang, John K. Hutchison, Rao Kotamarthi, Jeremy Feinstein, Haiwen Guan, Romit Maulik, Vijay P. Ramalingam, Jason Stock, Tom Wall
arXiv:2609.24631v1 Announce Type: cross
Abstract: Autonomous parking in nonconvex and narrow environments remains challenging. Although optimal-control methods can explicitly enforce vehicle dynamics...
By Zhengbao Yao, Yuanfu Luo, Kehan Xue
KwaiMind is a commercial image editing system that combines general editing capabilities with e-commerce specialization. It uses an agent-based data engine with 1.8 million editing pairs and a multimodal diffusion transformer trained through pre‑training, fine‑tuning, preference optimization, and online reinforcement learning. The system is guided by a vision‑language judge and specialized rewards for click‑through rate, text rendering, and product consistency, and it achieves top scores on ImgEdit, GEdit, REDEdit, and the new Ecom‑Bench, while improving predicted and actual CTR in offline and online experiments.
By Junlong Wu, Zijun Li, Yuting Hu, Jia Sun, Pengcheng Wei, Yimin Zhou, Honglie Wang, Huaiqing Wang, Dewen Fan, Fei Zuo, Haixuan Gao, Lihui Peng, Tingxuan She, Yuqing Li, Boheng Zhang, Fan Yang, Wenwu Ou
arXiv:2609.17475v2 Announce Type: replace
Abstract: Local agents need memory for model execution and working history. We present JustFit, an MLX runtime that coordinates their overlapping allocations...
By Yuhua Chen
PERSONAWEAVER is a new approach to procedural character generation that separates world building from behavioral specification, using manually curated banks of moral positions and conversational reactions to diversify character behavior. By applying this method across ten realistic and fantastical settings and three large language models, the system produces broader moral and interactional response distributions, varied interpersonal language, response length, sentiment, and less archetypal world attribute combinations compared to prior work.
By Maan Qraitem, Kate Saenko, Bryan A. Plummer
Vorch-Human is a unified framework for human‑centric audio‑visual generation that handles multiple tasks—animating a person from speech, jointly generating speech and video from a voice reference, and synthesizing a scene from paired appearance and voice references—using a single dual‑stream audio‑video diffusion transformer. The model incorporates clean condition‑audio and condition‑video tokens, per‑token task embeddings, temporal position types, condition masks, and a shared multimodal prompt encoder to express diverse inputs such as driving speech, timbre examples, first frames, and subject images. A two‑level data pipeline supplies the necessary supervision by extracting speech, appearance, and timbre annotations from clips and linking consistent identity and outfit references across videos, while a frozen‑prefix recurrence enables long‑form audio‑driven generation with reduced boundary discontinuity and identity drift.
By Yang Ding, Haoran Yu, Xin Ma, Yulei Lu, Menglin Han, Yaole Wang, Siqian Yang, Gang Yue, Kaihao Zhang, Yaohui Wang, Lin Ma
arXiv:2609.23997v1 Announce Type: cross
Abstract: Multi-robot collaboration could enable more efficient and scalable solutions to complex robotic tasks, but collaboration under partial observability...
By Dorian Benhamou Goldfajn, Mason Nakamura, Saaduddin Mahmud, Justin Svegliato, Kyle H. Wray, Shlomo Zilberstein
arXiv:2609.24526v2 Announce Type: replace
Abstract: Physical AI requires models to ground visual and linguistic understanding in real-world environments while accounting for environmental constraints...
By Foundation Model, Li Auto Inc
arXiv:2605.00915v2 Announce Type: replace
Abstract: Probing frozen vision transformers typically uses permutation-invariant aggregation (GAP or $\texttt{[CLS]}$), treating patch tokens as an unstruct...
By Zice Wang, Zhenyu Zhang
SF October 14th: A Birds of a Feather Session on Agentic Engineering
I'm hosting an evening event with Jesse Vincent in San Francisco on Wednesday 14th October for people who are building weird and in...
Transformer language models process sequences token by token in an autoregressive manner, making growing contexts increasingly expensive. Yet many adjacent token spans are highly predictable or freque...
ChatGPT Ads is expanding to Southeast Asia and Taiwan, giving eligible businesses new ways to reach people across more than 60 countries.
Objective assessment of robotic surgery uses instrument kinematics, which must be reconstructed when only video is available. We introduce a kinematic reconstruction network for estimating instrument...
Text-to-speech systems increasingly process user-generated text (UGT) such as ppl and imo, whose pronunciation must be inferred from the canonical rather than surface form. We introduce UGTPhon, the f...
Learn how Airbnb is expanding access to GPT-6 Astra and OpenAI frontier models to help engineering teams solve bugs, design systems, and ship faster.
The article reports the release of new AI models: Claude Opus 5.5 by Anthropic and GPT‑6 Sol and GPT‑6 Luna by OpenAI, noting that GPT‑6 variants are priced at half the cost of their GPT‑5.6 counterparts. It provides a detailed pricing table comparing input, cached input, and output costs across several models, highlighting how GPT‑6 Luna is among the cheapest ever offered by OpenAI. The author also comments on visual differences in model outputs, noting that GPT‑6 outputs are more muted compared to GPT‑5.6.
Multi-agent debate (MAD) has emerged as a promising paradigm for improving the reasoning accuracy of large language models (LLMs) through iterative peer interaction. Communication topology plays a cen...
Surgical procedures follow a phase-to-step hierarchy, yet the video-language models used to recognize them are evaluated with flat per-level metrics that ignore cross-level coherence and error structu...
The OpenAI Blog announces that GPT‑6 enhances prompt caching, achieving higher cache hit rates and introducing new diagnostics, explicit breakpoints, and controls. These features are designed to reduce latency and costs for users. The post highlights the technical improvements that make GPT‑6 more efficient and cost‑effective.