Category intelligence

Research Briefing — June 17, 2026

627 current items analyzed and ranked.

Executive synthesis

Research Summary

Today's research is led by major-lab model reports emphasizing efficiency, alongside safety/alignment methodology and rigorous evaluation.

Models & Architectures

Nemotron 3 Ultra (NVIDIA) is a 550B-total/55B-active MoE hybrid Mamba-Attention model trained on 20T tokens with 1M context and ~6x higher throughput, targeting agentic reasoning.
Ling and Ring 2.6 scale efficient agentic intelligence to trillion-parameter families via architectural-migration pretraining.
Rethinking the Role of Efficient Attention offers systematic mechanistic and scaling insights into hybrid full-attention plus sliding-window/recurrent designs.
VibeThinker-3B pushes verifiable reasoning toward frontier performance in a strict small-model regime via Spectrum-to-Signal training.

Safety & Alignment

Predicting LLM Safety Before Release forecasts deployment behavior by replaying prior conversations in a privacy-preserving simulation.
Synthetic document finetuning (DeepMind) instills positive traits in Gemini 3 Flash through synthetic-document midtraining.

Data & Evaluation

Spokes directly optimizes set-level pretraining diversity using the G-Vendi score with exponentiated gradient descent.
How Inference Compute Shapes Frontier LLM Evaluation shows inference-scaling changes rankings across 12 frontier models on hard benchmarks.
First Proof Second Batch has prominent mathematicians evaluate frontier AI on ten research-level math problems.
ProCUA-SFT releases a 3.1M-sample synthetic dataset to address the computer-use agent training bottleneck.

Key Themes

Language Models · 25Efficiency and Architectures · 6LLM Agents and Memory · 10Reinforcement Learning and Reasoning · 11Benchmarks and Evaluation · 34Evaluation and Benchmarks · 12Agents and Self-Evolving Systems · 20AI Safety, Bias, and Alignment · 9AI Safety & Alignment · 7Reasoning and Interpretability · 9

Primary evidence

Top Ranked Signals

Research arXiv (Computation and Language) Jun 17

Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning

By NVIDIA (Allan), : (Allan), Aaron Blakeman (Allan), Aaron Thomas (Allan), Aastha Jhunjhunwala (Allan), Abhibha Gupta (Allan), Abhinav Khattar (Allan), Adam Rajfer (Allan), Adi Renduchintala (Allan), Adil Asif (Allan), Aditya Vavre (Allan), Adriana Flores Miranda (Allan), Ahmad Bilal (Allan), Aileen Zaman (Allan), Ajay Hotchandani (Allan), Akanksha Shukla (Allan), Akhiad Bercovich (Allan), Aleksander Ficek (Allan), Alex Gronskiy (Allan), Alex Kondratenko (Allan), Alex Steiner (Allan), Alex Ye (Allan), Alexander Bukharin (Allan), Alexandre Milesi (Allan), Ali Taghibakhshi (Allan), Alice Gatti (Allan), Alisa Liu (Allan), Alok Kumar (Allan), Amar Phanishayee (Allan), Ameya Sunil Mahabaleshwarkar (Allan), Amir Klein (Allan), Amit Zuker (Allan), Amnon Geifman (Allan), Anahita Bhiwandiwalla (Allan), Ananth Subramaniam (Allan), Andrea Santilli (Allan), Andrew Fulks (Allan), Andrew McHarg (Allan), Andrew Tao (Allan), Andrii Skliar (Allan), Anjulie Agrusa (Allan), Ankur Srivastava (Allan), Ankur Verma (Allan), Anna Shors (Allan), Anna Warno (Allan), Antoni-Joan Solergibert I Llaquet (Allan), Arham Mehta (Allan), Arkadiusz Nowaczynski (Allan), Arti Jain (Allan), Ashwath Aithal (Allan), Ashwin Poojary (Allan), Asif Ahamed (Allan), Asit Mishra (Allan), Asma Kuriparambil Thekkumpate (Allan), Atefeh Sohrabizadeh (Allan), Avinash Kaur (Allan), Avinash Vem (Allan), Ayush Dattagupta (Allan), Barath Subramaniam Anandan (Allan), Bardiya Sadeghi (Allan), Ben Lanir (Allan), Benedikt Schifferer (Allan), Besmira Nushi (Allan), Bilal Kartal (Allan), Bill Thiede (Allan), Bita Darvish Rouhani (Allan), Bo Deng (Allan), Bob Schatz (Allan), Boris Ginsburg (Allan), Boxin Wang (Allan), Brad Nemire (Allan), Brandon Norick (Allan), Brian Dang (Allan), Brian Westphal (Allan), Brian Yu (Allan), Brucek Khailany (Allan), Bryan Catanzaro (Allan), Carlo del Mundo (Allan), Caryln Aarish (Allan), Chankyu Lee (Allan), Chantal Hwang (Allan), Charbel Sakr (Allan), Charles Wang (Allan), Charlie Truong (Allan), Chen Cui (Allan), Cheng Cheng (Allan), Cheng-Ping Hsieh (Allan), Chenghao Zhang (Allan), Chenhui Deng (Allan), Chintan Patel (Allan), Chris Alexiuk (Allan), Christian Cosgrove (Allan), Christian Munley (Allan), Christine Harvey (Allan), Christopher Parisien (Allan), Chunyang Shen (Allan), Coco Li (Allan), Collin Neale (Allan), Cynthia Gao (Allan), Cyril Meurillon (Allan), Dan Gil (Allan), Dan Su (Allan), Dan Zhao (Allan), Dane Corneil (Allan), Daniel Afrimi (Allan), Daniel Egert (Allan), Daniel Korzekwa (Allan), Daniel Lo (Allan), Daniel Machlab (Allan), Daniel Serebrenik (Allan), Daniil Sorokin (Allan), Daria Gitman (Allan), Daria Levy (Allan), Darko Stosic (Allan), David Mosallanezhad (Allan), David Yu (Allan), Davit Karamyan (Allan), Deena Donia (Allan), Deep Debroy (Allan), Deepak Narayanan (Allan), Devin O'Kelly (Allan), Dheeraj Peri (Allan), Dhruv Nathawani (Allan), Di (Allan), Wu, Dima Rekesh, Divyanshu Kakwani, Donald Plummer, Dong Anh, Dongfeng Yu, Dongfu Jiang, Donnie Kim, Dorrin Poorkay, Duncan Riach, Dusan Stosic, Dustin VanStee, Eavan Meng, Edgar Minasyan, Edward Lin, Eileen Margaret Peters Long, Elad Sarafin, Elad Segal, Elena Lantz, Ellie Evans, Elliott Ning, Eric Chung, Eric Harper, Eric Pham-Hung, Eric Tramel, Eric Yang, Erick Galinkin, Erik Pounds, Erika Goncalves Goncalves, Evan Briones, Evan Wu, Evelina Bakhturina, Evgeny Tsykunov, Ewa Dobrowolska, Faisal Ladhak, Farzan Memarian, Fay Wang, Fei Jia, Felipe Soares, Felipe Vieira Frujeri, Feng Chen, Fengguang Lin, Ferenc Galko, Frank Sun, Frankie Siino, Frida Hou, Gal Hubara Agam, Gal Kaplun, Gantavya Bhatt, Gargi Prasad, Garvit Kulshreshtha, George Armstrong, Gerald Shen, Giulio Borghesi, Gordana Neskovic, Gorkem Batmaz, Grace Lam, Greg Mason, Greg Pauloski, Grigor Nalbandyan, Grzegorz Chlebus, Grzegorz Karch, Guan-Ting Liu, Guoming Zhang, Guyue Huang, Haggai Maron, Haifeng Qian, Haim Elisha, Haoxing Ren, Haran Kumar Shiv Kumar, Haribhau Hud, Harris Nover, Harrison Saturley Hall, Hayate Iso, Helen Ngo, Herbert Hum, Herman Sahota, Hexin Wang, Himanshu Soni, Hovhannes Tamoyan, Hua Li, Huanhuan Chen, Hui Li, Hui Wang, Huy Nguyen, Ian Chiles, Ido Galil, Ido Shahaf, Igor Gitman, Igor Shovkun, Ilya Loshchilov, Ingo Guehring, Itamar Schen, Itay Levy, Itay Neeman, Ivan Moshkov, Izik Golan, Izzy Putterman, Jaemin Choi, Jakub Slowikowski, Jan Kautz, Jane Polak Scowcroft, Jared Casper, Jatin Mitra, Jeffrey Glick, Jenny Chen, Jesse Oliver, Jiacheng Xu, Jiafan Zhu, Jialin Song, Jian Zhang, Jiantao Jiao, Jiaqi Zeng, Jie Lou, Jim King, Jimmy Zhang, Jingquan Wang, Jinhang Choi, Jinju Chu, Joey Conway, Joey Guman, Johan Jatko, Johannes Rausch, John Kamalu, John Roberts, Johnny Greco, Johnny Mensel, Jonah Alben, Jonas Yang, Jonathan Cohen, Jonathan Raiman, Joseph Jennings, Joshua Mabry, Joshua Pierce, Joyjit Daw, Julien Veron Vialard, Junkeun Yi, Jupinder Parmar, Kajal Jain, Kan Zhu, Kari Briski, Katherine Cheung, Katherine Luna, Keith Willowhawk, Keith Wyss, Keshav Santhanam, Kevin Shih, Kezhi Kong, Khanh Nguyen, Khushi Bhardwaj, Kirthi Shankar Sivamani, Konstantinos Krommydas, Krishna C. Puvvada, Krzysztof Pawelec, Kumar Anik, Kyle Keprios, Kylie Day, Lawrence McAfee, Leo Du, Leon Derczynski, Li Ding, Linda Liu, Lingjie Wu, Lior Kadoch, Lizzie Wei, Luis Vega, Luke Robison, Lun Su, Maarten Van Segbroeck, Maciej Jakub Mikulski, Maer Rodrigues de Melo, Magda Sypula, Mahan Fathi, Makesh Narsimhan Sreedhar, Makesh Tarun Chandran, Manoj Kilaru, Maor Ashkenazi, Marc Cuevas, Marc Romeijn, Marcin Chochowski, Mark Cai, Mark Mozolewski, Markus Kliegl, Marta Stepniewska-Dziubinska, Martyna Patelka, Mattei Machczynski, Matvei Novikov, Mauricio Ferrato, Maximilian Golub, Mehrzad Samadi, Melissa Corpuz, Mengru Wang, Mengxi Wu, Meredith Price, Meriem Boubdir, Micah Schaffer, Michael Andersch, Michael Boone, Michael Gschwind, Michael Lightstone, Michael Loh, Michal Bien, Michal Zawalski, Michelle Gill, Miguel Martinez, Mikail Khona, Mike Chrzanowski, Mike Houston, Mingyuan Ma, Minseok Lee, Mohamed Fawzy, Mohammad Dabbah, Mohammad Shoeybi, Mostofa Patwary, Nabin Mulepati, Najeeb Nabwani, Namit Dhameja, Narimane Hennouni, Natalie Hereth, Nathaniel Pinckney, Nave Algarici, Nave Assaf, Netanel Haber, Nicholas Knight, Nick Reamaroon, Nickson Quak, Nidhi Bhatia, Nikhil Desai, Nikolai Ludwig, Nima Tajbakhsh, Ning Xu, Nir Ailon, Nirmal Juluru, Nitin Nitin, Ofri Masad, Oleg Rybakov, Oleksii Hrinchuk, Oleksii Kuchaiev, Olivia Viessmann, Olivier Delalleau, Oluwatobi Olabiyi, Omer Ullman Argov, Omri Puny, Oren Tropp, Pablo Ribalta, Pallab Bhattacharya, Panos Lampropoulos, Parth Mannan, Pasha Shamis, Patrick Legresley, Paul Gibbons, Pavlo Molchanov, Pawel Morkisz, Peter Dykas, Peter Jin, Pierre-Yves Aquilanti, Pinky Xu, Piotr Januszewski, Piotr Laskiewicz, Pooya Jannaty, Prakash Gurumurthy, Pranav Prashant Thombre, Prasoon Varshney, Pritam Gundecha, Przemek Tredak, Puhui Meng, Qiyu Wan, Rabeeh Karimi Mahabadi, Rachel Oberman, Rachit Garg, Radha Sri-Tharan, Rahul Kandu, Rakshit Sanadhya, Ran El-Yaniv, Ran Zilberstein, Rasoul Shafipour, Ray Macalisang, Rayen Tian, Reka Kovacs, Renjie Pi, Rick Izzo, Rima Shahbazyan, Rishabh Garg, Rishi Puri, Rita Fernandes Neves, Ritchie Zhao, Ritika Borkar, Ritu Gala, Riyad Islam, Robert Clark, Robert Hesse, Robert Kirby, Roger Waleffe, Rohit Watve, Roi Koren, Ron Banner, Ruoxi Zhang, Russell J. Hewett, Ryan Prenger, Ryan Stewart, Ryota Egashira, Sadegh Mahdavi, Saee Paliwal, Sagar Singh, Sahil Modi, Salika Dave, Samantha Shinagawa, Samuel Kriman, Sandip Bhaskar, Sangkug Lym, Sanjay Kariyappa, Sanjeev Satheesh, Saran Vikas Murari, Satish Pasumarthi, Saurabh Mishra, Saurav Muralidharan, Scott Hara, Sean Narentharen, Selvaraj Anandaraj, Seonjin Na, Seonmeyong Bak, Seonmyeong Bak, Sepehr Sameni, Seph Mard, Serge Panev, Seth Henneman, Seth Poulos, Shahar Mor, Shantanu Acharya, Shaona Ghosh, Sharath Turuvekere Sreenivas, Sharon Mendelson, Shaun Kotek, Shawn Wang, Shay Aharon, Shaya Gharghabi, Sheng-Chieh Lin, Shi Chen, Shiqing Fan, Shirish Baskaran, Shreya Gopa, Shrimai Prabhumoye, Shubham Pachori, Shubham Toshniwal, Shuoyang Ding, Shwetha Krishnamurthy, Siddharth Singh, Simeng Sun, Sirshak Das, Sivakumar Arayandi Thottakara, Smita Ithape, Somshubra Majumdar, Soumye Singhal, Sri Harsha Singudasu, Sridhar Bhuvanapalli, Srimukh Veccham, Stas Sergienko, Stefania Alborghetti, Stephen Ge, Su Rong, Sugam Dipak Devare, Sukrit Rao, Sumeet Kumar Barua, Sungsoo Ha, Sunny Gai, Suriya Gunasekar, Suseella Panguluri, Suyog Gupta, Sviataslau Hinzburh, Sweta Priyadarshi, Syeda Nahida Akter, Talor Abramovich, Tan Bui, Tanay Varshney, Tatevik Ter-Hovhannisyan, Teodor-Dumitru Ene, Terry Kong, Thanh Do, Tianhe Zhang, Tiffany Moore, Tijmen Blankevoort, Tim Moon, Tiyasa Mitra, Tom Balough, Tomasz Grzegorzek, Tomasz Hliwiak, Tomer Asida, Tomer Bar Natan, Tomer Keren, Tomer Ronen, Tony Salim, Tony Wang, Traian Rebedea, Tugrul Konuk, Twinkle Vashishth, Udi Karpas, Ushnish De, Vahid Noorozi, Venkat Srinivasan, Venmugil Elango, Vibhor Agrawal, Victor Cui, Vijay Korthikanti, Vikas Mehta, Vinay Rao, Virginia Wu, Vitaly Kurin, Vitaly Lavrukhin, Vladimir Anisimov, Vu Pham, Wanli Jiang, Wasi Uddin Ahmad, Wataru Ishihara, Wei Du, Wei Ping, Weiheng Chai, Wenliang Dai, Wesley Helmholz, Will Jennings, Will Zhu, Wojciech Prazuch, Xiaowei Ren, Xiwen Yu, Yan Breek, Yang Chen, Yang Yu, Yangyi Chen, Yaniv Galron, Yashaswi Karnati, Yejin Choi, Yev Meyer, Yi-Fu Wu, Yian Zhang, Ying Lin, Yonatan Geifman, Yonggan Fu, Youngeun Kwon, Yu Yao, Yugi Guvvla, Yuki Huang, Yunsheng Liu, Zach Moshe, Zachary Newell, Zhilin Wang, Zhiyu Li, Zhongbo Zhu, Zhuolin Yang, Zihan Liu, Zijie Yan, Zsolt-Alon Wertheimer

43 score

AI Analysis

NVIDIA introduces Nemotron 3 Ultra, a 550B-total/55B-active MoE hybrid Mamba-Attention model pretrained on 20T tokens with 1M context, achieving ~6x higher inference throughput than SOTA public LLMs at comparable accuracy. It combines LatentMoE, multi-token prediction, NVFP4 pretraining, multi-environment RLVR, and on-policy distillation for efficient agentic reasoning.

arXiv:2606.15007v1 Announce Type: new Abstract: We introduce Nemotron 3 Ultra, a 550 billion total and 55 billion active parameter Mixture-of-Experts Hybrid Mamba-Attention language model. We pre-trained Nemotron 3 Ultra on 20 trillion text tokens, then extended the context length to 1M tokens, and post-trained using Supervised Fine Tuning (SFT), Reinforcement Learning (RL), and Multi-teacher On-Policy Distillation (MOPD). Nemotron 3 Ultra is our most capable model yet, employing multiple key t

Language ModelsMixture-of-ExpertsState-Space ModelsEfficiencyAgentic Reasoning

Research LessWrong Jun 16

Predicting LLM Safety Before Release by Simulating Deployment

By Tomek Korbak

78 score

AI Analysis

Describes a deployment simulation method for forecasting how a new model will behave before release by replaying prior conversations in a privacy-preserving way with the candidate model. In a GPT-5.4 study it predicted the direction of behavior change 92% of the time for categories shifting by 1.5x or more, far above a 54% baseline.

Paper linkBefore releasing a new model, labs need to understand not just what it can do, but how it is likely to behave in real-world use, including where it might introduce new risks. This becomes even more important as capabilities increase. As part of our pre-deployment safety review, we leverage targeted evaluations, red-teaming, and other checks to understand model behavior. We’ve now started using a method for simulating model deployments before they happen, which adds a complementary sign

AI SafetyEvaluationPre-deployment TestingLanguage Models

Research arXiv (Computation and Language) Jun 17

Ling and Ring 2.6 Technical Report: Efficient and Instant Agentic Intelligence at Trillion-Parameter Scale

By Ang Li, Ben Liu, Bin Han, Bin Hu, Bin Jing, Binbin Hu, Bing Li, Cai Chen, Caizhi Tang, Changxin Tian, Chao Huang, Chao Zhang, Chen Liang, Chen Qian, Chengfu Tang, Chengyao Wen, Chilin Fu, Chunwei Wu, Cong Zhang, Cunyin Peng, Daixin Wang, Dalong Zhang, Deng Zhao, Dingnan Jin, Dingyuan Zhu, Donghao Zhang, Fan Yuan, Fangzheng Zhao, Fanzhuang Meng, Feifan Wu, Feng Xu, Fengbin Fang, Gangshan Wang, Guodong Yang, Hailin Zhao, Haitao Wang, Haitao Zhang, Hanxiao Zhang, Hanzi Wang, Hao Dai, Hao Liu, Hao Qian, Hao Wu, Haoxiong Liu, Haoyu Xu, Heng Zhang, Hong Liu, Hongliang Zhang, Hongrui Liu, Hongxun Li, Hongzhi Ruan, Huaidong Xiong, Huihuang Zheng, Huikang Tang, Jia Guo, Jia Li, Jia Liu, Jiameng Wang, Jiaming Liu, Jiannan Shi, Jianping Wei, Jiaolong Yang, Jiapeng Wang, Jie Gao, Jie Wang, Jiewei Wu, Jin Yang, Jinjin Li, Jinjing Huang, Jinquan Sun, Jinyao Chen, Juanhui Tu, Jun Liu, Jun Mei, Jun Xu, Jun Zhou, Junjie Ou, Junnan Sipan, Junpeng Fang, Kaihong Zhang, Kaiqin Hu, Ke Shi, Kuan Xu, Kun Tang, Kunlong Chen, Lanyin Mei, Lei Chen, Lei Liang, Lei Xu, Li Tang, Liang Jiang, Liangcheng Fu, Lihui Zhang, Linfeng Shi, Lintao Ma, Liyuan Liu, Longfei Li, Longfei Zheng, Lu Liu, Lu Yu, Man Li, Meiqi Zhu, Meng Li, Mengjie Gao, Mengshu Sun, Mingming Yin, Mingyang Zhang, Mingyuan Fan, Nuo Xu, Pan Tang, Peijie Jiang, Peilong Zhao, Peng Lin, Pingping Liu, Qi Zuo, Qian Zhao, Qiang Cheng, Qianggang Cao, Qiaoben Bao, Qing Cui, Qingyuan Yang, Qitao Shi, Qiyin Huang, Qizheng Zhou, Quan Wan, Runyuan Zhao, Shaomian Zheng, Shaowei Wei, Shengnan Zhang, Shuaicheng Li, Shujie Li, Shuo Zhang, Sikang Bian, Tianchu Yao, Tiange Xu, Tianshu Wang, Ting Guo, Tinghao Wang, Tingwei Huang, Tong Zhao, Tongkai Yang, Wang Hong, Wanli Gu, Wei Lu, Weichang Wu, Weiguang Han, Weiquan Li, Wenbo Shen, Wenjing Fang, Wenzhi Tang, Xiang Shu, Xiao Shi, Xiaodong Yan, Xiaolu Zhang, Xiaopei Wan, Xiaqing Sun, Xin Zhao, Xingyu Lu, Xinxing Yang, Xinyao Tang, Xinyu Kong, Xinyu Liu, Xiong Xu, Xuan Sun, Xudong Han, Xudong Wang, Xujie Shen, Yalin Zhang, Yangyang Hou, Yankun Ren, Yao Zhao, Ye Chen, Yeyang Chen, Yibo Cao, Yifan Zuo, Yijie Chen, Ying Li, Yingjie Song, Yingxue Li, Yiqi Wang, Yixuan Sun, Yizhu Xiao, Yongfei Xu, Yu Liu, Yuchen Fang, Yue Gao, Yue Yu, Yue Zhang, Yuqi Zhang, Yuxiao He, Yuxiao Lu, Yuxin Tian, Yuxuan Li, Yuzhuo Fu, Zhankai Xu, Zhaoxin Huan, Zhenduo Zhang, Zhengke Gui, Zhengyu Huang, Zhenjun Ma, Zhenxuan Pan, Zheping Qu, Zhibo Zhu, Zhidong Fan, Zhigang Huangfu, Zhihao Wang, Zhiqiang Zhang, Zhizhen Liu, Zhuyan Zhou, Zibin Lin, Zihang Zeng, Zihao Wang, Zilong Wang, Ziqi Liu, Zitao Xuan, Zixuan Cheng, Zujie Wen, Zuoli Tang

37 score

AI Analysis

Technical report on Ling-2.6 and Ring-2.6, trillion-parameter model families upgraded from Ling-2.0 via architectural-migration pretraining and large-scale post-training, with Ling optimized for low-latency response and Ring for deep reasoning and agentic workflows. It emphasizes co-design of architecture, objectives, serving, and agents for efficient deployment.

arXiv:2606.15079v1 Announce Type: new Abstract: Efficient and scalable agentic intelligence requires models that can deliver both low-latency responses and strong reasoning capabilities while remaining practical to train, serve, and deploy. In this report, we present Ling-2.6 and Ring-2.6, a family of models designed to address this challenge at scale. Ling-2.6 is optimized for instant response generation and high capability per output token, whereas Ring-2.6 is tailored for deeper reasoning an

Language ModelsMixture-of-ExpertsAgentic ReasoningEfficiency

Research LessWrong Jun 15

Synthetic document finetuning for instilling positive traits

By CallumMcDougall

75 score

AI Analysis

A DeepMind interpretability team research update on instilling positive traits in Gemini 3 Flash via synthetic document midtraining followed by synthetic chat finetuning, building on Marks et al and Li et al. They report the chat finetuning robustly instills traits that generalize out-of-distribution and share practical takeaways for improving effectiveness.

This is the fifth in a series of informal research updates from the Google DeepMind Language Model Interpretability team, in interpretability and adjacent areas. The fourth post can be found here.TLDR: Via adapting the methods of Marks et al and Li et al, we train Gemini 3 Flash to have certain traits/values by midtraining it on documents about how Gemini has those properties, followed by finetuning it on synthetic chat data where it demonstrates those properties. The chat finetuning is effectiv

AlignmentInterpretabilitySynthetic DataLanguage ModelsFine-tuning

Research arXiv (Machine Learning) Jun 17

ProCUA-SFT Technical Report

By Jaehun Jung, Ximing Lu, Brandon Cui, Muhammad Khalifa, Shaokun Zhang, Hao Zhang, Jin Xu, Amala Sanjay Deshmukh, Karan Sapra, Andrew Tao, Yejin Choi, Jan Kautz, Mingjie Liu, Yi Dong

72 score

AI Analysis

ProCUA-SFT introduces a 3.1M-sample synthetic dataset for training computer-use agents, generated through an automated pipeline that synthesizes grounded desktop tasks. It addresses the negative transfer problem where the largest public dataset (AgentNet) actually degrades agent performance during fine-tuning.

arXiv:2606.17321v1 Announce Type: new Abstract: Training computer-use agents (CUAs) -- models that interact with graphical desktops through screenshots and keyboard/mouse actions -- requires large-scale, diverse trajectory data collected in full desktop environments. The largest public resource, AgentNet (22.5K human trajectories), leads to negative transfer when used for supervised fine-tuning (SFT): continuing training UI-TARS 7B on AgentNet causes OSWorld success rate to fall from 26.3% to 8

AgentsLanguage ModelsSynthetic Data

Research arXiv (Artificial Intelligence) Jun 17

First Proof Second Batch

By Mohammed Abouzaid, Nikhil Srivastava, Rachel Ward, Lauren Williams

70 score

AI Analysis

Tests several AI systems on ten research-level mathematics problems arising naturally in working mathematicians' research, providing problems, methodology, human and AI solutions, and referee reports. Assesses current AI's ability to solve genuine open-ended math research.

arXiv:2606.18119v1 Announce Type: new Abstract: To assess the ability of current AI systems to correctly solve research-level mathematics problems, we tested several AI systems on a set of ten problems in a broad range of mathematical fields; these problems arose naturally in the research process of the contributors. This document includes the problems, our methodology, and the results of our testing. We provide links to supplementary documents including the human solutions, the AI-generated so

AI for MathEvaluationReasoningFrontier Models

Research arXiv (Computation and Language) Jun 17

Spokes: Optimizing for Diverse Pretraining Data Selection

By Clarence Lee, Yejin Choi, Luke Zettlemoyer, Pang Wei Koh, Hai Leong Chieu

70 score

AI Analysis

Introduces Spokes, a probabilistic diversification framework using the G-Vendi score optimized via exponentiated gradient descent to directly optimize set-level diversity in pretraining data selection. It produces substantially more diverse subsets than random sampling, improving performance under fixed data budgets.

arXiv:2606.15216v1 Announce Type: new Abstract: Diversity plays a critical role in data selection, improving performance under fixed data budgets by reducing redundancy and repetition. However, optimizing for diversity is inherently challenging, as it is a set-level property that depends on interactions between data points rather than individual examples. As a result, existing approaches typically rely on proxies or approximations, which often fail to ensure sufficiently diverse subsets. In thi

Data SelectionPretrainingLanguage ModelsOptimization

Research arXiv (Computation and Language) Jun 17

Rethinking the Role of Efficient Attention in Hybrid Architectures

By Ziqing Qiao, Yinuo Xu, Chaojun Xiao, Zhou Su, Zihan Zhou, Yingfa Chen, Xiaoyue Xu, Xu Han, Zhiyuan Liu

70 score

AI Analysis

Conducts a systematic analysis of hybrid LLM architectures combining full attention with efficient modules (sliding-window attention, recurrent mixers), examining scaling behavior, mechanism, and design. It finds efficient-attention design mainly affects how fast long-context capability emerges while hybrids converge to similar performance with enough training.

arXiv:2606.15378v1 Announce Type: new Abstract: Modern language models increasingly adopt hybrid architectures that combine full attention with efficient attention modules, such as sliding-window attention (SWA) and recurrent sequence mixers. However, how these efficient modules shape model capabilities remains poorly understood. To address this gap, we conduct a systematic analysis across hybrid architectures from three perspectives: scaling behavior, mechanism analysis, and architecture desig

Neural ArchitecturesEfficient AttentionLong ContextLanguage Models

Research arXiv (Computation and Language) Jun 17

VibeThinker-3B: Exploring the Frontier of Verifiable Reasoning in Small Language Models

By Sen Xu, Shixi Liu, Wei Wang, Jixin Min, Yingwei Dai, Zhibin Yin, Yirong Chen, Xin Zhou, Junlin Zhang

35 score

AI Analysis

VibeThinker-3B is a compact 3B-parameter dense model that pushes verifiable reasoning within a strictly small-model regime using a Spectrum-to-Signal post-training pipeline of curriculum SFT, multi-domain RL, and offline self-distillation. It reports strong results on AIME26 (94.3), LiveCodeBench v6 (80.2 Pass@1), and out-of-distribution tasks. Notable for showing frontier reasoning achievable at small scale.

arXiv:2606.16140v1 Announce Type: cross Abstract: This technical report introduces VibeThinker-3B, a compact dense model with 3B parameters developed to investigate how far verifiable reasoning can be pushed within a strictly small-model regime. Building upon the Spectrum-to-Signal post-training paradigm, we systematically enhance the model through an optimized pipeline that includes curriculum-based supervised fine-tuning, multi-domain reinforcement learning, and offline self-distillation. Exp

Small Language ModelsReasoningReinforcement LearningPost-Training

Research arXiv (Artificial Intelligence) Jun 17

Nothing from Something: Can a Language Model Discover 0?

By Phoebe Zeng, Thomas L. Griffiths, Brenden M. Lake

68 score

AI Analysis

Investigates whether language models can independently discover the mathematical concept of zero, using arithmetic as a case study for out-of-distribution mathematical generalization. Frames mathematical discovery as requiring strong novel-structure hypothesis generation.

arXiv:2606.17289v1 Announce Type: new Abstract: AI systems based on artificial neural networks are being developed with aspirations of pushing the boundary of human mathematical knowledge. A key question for these systems is how much they can reach beyond their training data. Mathematical discovery requires a strong form of out of distribution generalization; the ability to hypothesize genuinely new - and potentially logically more powerful - mathematical structures. It has been hypothesized th

Language ModelsMathematical ReasoningCognitive ScienceGeneralization

Research arXiv (Artificial Intelligence) Jun 17

How Inference Compute Shapes Frontier LLM Evaluation

By Jessica McFadyen, Ole Jorgensen, Harry Coppock, Kevin Wei, Cozmin Ududec

68 score

AI Analysis

Evaluates up to 12 frontier LLMs on seven hard benchmarks (software engineering, math, medicine, cybersecurity) under controlled inference-scaling interventions (token budgets, context compaction, repeated sampling). Shows that single restrictive compute budgets can understate true model capability.

arXiv:2606.17930v1 Announce Type: new Abstract: AI evaluations are shifting toward harder tasks that benefit from longer trajectories involving tool use and iterative problem solving. As a result, performance is increasingly sensitive to the amount and allocation of compute available at test time ("inference compute"). Yet many evaluations still report performance at a single restrictive budget, meaning that low scores may reflect the evaluation setup rather than the model's underlying capabili

EvaluationInference ComputeTest-Time ScalingFrontier Models

Research arXiv (Artificial Intelligence) Jun 17

Vision-language models for chest radiography do not always need the image

By Mahshad Lotfinia, Sebastian Ziegelmayer, Lisa Adams, Daniel Truhn, Andreas Maier, Soroosh Tayebi Arasteh

68 score

AI Analysis

Introduces a causal audit that intervenes on chest X-ray images to test whether vision-language models actually use the image rather than exploiting finding-name priors. Strikingly, a text-only model comes within 5.7 accuracy points of the best multimodal model, exposing a fundamental evaluation flaw.

arXiv:2606.17710v1 Announce Type: cross Abstract: Medical vision-language models report strong chest radiograph accuracy, and this is increasingly read as evidence that they use the image. That inference is unsafe: a model exploiting finding-name priors scores like one that reads the scan, and no standard benchmark separates them. We introduce a causal audit that intervenes on the image, occluding the relevant region, occluding an irrelevant one, and swapping in another patient's same-label sca

Vision-Language ModelsMedical ImagingEvaluationAI Safety