Tootfinder

Tencent open sources translation models Hunyuan-MT-7B and Hunyuan-MT-Chimera-7B, which support 33 languages, claiming they beat established models in benchmarks (Jonathan Kemper/The Decoder)
https://the-decoder.com/tencent-open-sources-two-high-performing…

Tencent open sources two high-performing translation models
Chinese tech giant Tencent has open sourced two specialized translation models, claiming they outperform established tools like Google Translate in international benchmarks.

Med-RewardBench: Benchmarking Reward Models and Judges for Medical Multimodal Large Language Models
Meidan Ding, Jipeng Zhang, Wenxuan Wang, Cheng-Yi Li, Wei-Chieh Fang, Hsin-Yu Wu, Haiqin Zhong, Wenting Chen, Linlin Shen
https://arxiv.org/abs/2508.21430

Med-RewardBench: Benchmarking Reward Models and Judges for Medical Multimodal Large Language Models
Multimodal large language models (MLLMs) hold significant potential in medical applications, including disease diagnosis and clinical decision-making. However, these tasks require highly accurate, context-sensitive, and professionally aligned responses, making reliable reward models and judges critical. Despite their importance, medical reward models (MRMs) and judges remain underexplored, with no dedicated benchmarks addressing clinical requirements. Existing benchmarks focus on general MLLM c…

The Complexity of Defining and Separating Fixpoint Formulae in Modal Logic
Jean Christoph Jung, J\k{e}drzej Ko{\l}odziejski
https://arxiv.org/abs/2509.24583 https://

The Complexity of Defining and Separating Fixpoint Formulae in Modal Logic
Modal separability for modal fixpoint formulae is the problem to decide for two given modal fixpoint formulae $φ,φ'$ whether there is a modal formula $ψ$ that separates them, in the sense that $φ\modelsψ$ and $ψ\models\negφ'$. We study modal separability and its special case modal definability over various classes of models, such as arbitrary models, finite models, trees, and models of bounded outdegree. Our main results are that modal separability is PSpace-complete over words, that is,…

Image-Difficulty-Aware Evaluation of Super-Resolution Models
Atakan Topaloglu, Ahmet Bilican, Cansu Korkmaz, A. Murat Tekalp
https://arxiv.org/abs/2509.26398 https://

Image-Difficulty-Aware Evaluation of Super-Resolution Models
Image super-resolution models are commonly evaluated by average scores (over some benchmark test sets), which fail to reflect the performance of these models on images of varying difficulty and that some models generate artifacts on certain difficult images, which is not reflected by the average scores. We propose difficulty-aware performance evaluation procedures to better differentiate between SISR models that produce visually different results on some images but yield close average performan…

Double Descent as a Lens for Sample Efficiency in Autoregressive vs. Discrete Diffusion Models
Ahmad Fraij, Sam Dauncey
https://arxiv.org/abs/2509.24974 https://

Double Descent as a Lens for Sample Efficiency in Autoregressive vs. Discrete Diffusion Models
Data scarcity drives the need for more sample-efficient large language models. In this work, we use the double descent phenomenon to holistically compare the sample efficiency of discrete diffusion and autoregressive models. We show that discrete diffusion models require larger capacity and more training epochs to escape their underparameterized regime and reach the interpolation threshold. In the strongly overparameterized regime, both models exhibit similar behavior, with neither exhibiting a…

An analysis of AI training datasets, compiled by The Atlantic, shows AI models were trained on hundreds of thousands of YouTube videos from news publishers (Andrew Deck/Nieman Lab)
https://www.niemanlab.org/2025/10/hundred…

Hundreds of thousands of videos from news publishers like The New York Times and Vox were used to train AI models
YouTube channels from major news publishers and creators were in video data sets used by Microsoft, Meta, Snap, Runway AI, and Bytedance.

AHELM: A Holistic Evaluation of Audio-Language Models
Tony Lee, Haoqin Tu, Chi Heem Wong, Zijun Wang, Siwei Yang, Yifan Mai, Yuyin Zhou, Cihang Xie, Percy Liang
https://arxiv.org/abs/2508.21376

AHELM: A Holistic Evaluation of Audio-Language Models
Evaluations of audio-language models (ALMs) -- multimodal models that take interleaved audio and text as input and output text -- are hindered by the lack of standardized benchmarks; most benchmarks measure only one or two capabilities and omit evaluative aspects such as fairness or safety. Furthermore, comparison across models is difficult as separate evaluations test a limited number of models and use different prompting methods and inference parameters. To address these shortfalls, we introd…

Despite the hype around large AI models, many companies like Meta are using small models for routine tasks, finding them more practical and cost-effective (Christopher Mims/Wall Street Journal)
https://www.wsj.com…

Challenges and Applications of Large Language Models: A Comparison of GPT and DeepSeek family of models
Shubham Sharma, Sneha Tuli, Narendra Badam
https://arxiv.org/abs/2508.21377

Challenges and Applications of Large Language Models: A Comparison of GPT and DeepSeek family of models
Large Language Models (LLMs) are transforming AI across industries, but their development and deployment remain complex. This survey reviews 16 key challenges in building and using LLMs and examines how these challenges are addressed by two state-of-the-art models with unique approaches: OpenAI's closed source GPT-4o (May 2024 update) and DeepSeek-V3-0324 (March 2025), a large open source Mixture-of-Experts model. Through this comparison, we showcase the trade-offs between closed source models …

DeepMind: video models like Veo 3 could become general purpose foundation models for vision, like LLMs for text, using zero-shot "chain-of-frames" reasoning (Simon Willison/Simon Willison's Weblog)
https://simonwillison.net/2025/Sep/27/v…

Video models are zero-shot learners and reasoners
Fascinating new paper from Google DeepMind which makes a very convincing case that their Veo 3 model - and generative video models in general - serve a similar role in …