Apache Software License 2.0
PulseAugur coverage of Apache Software License 2.0 — every cluster mentioning Apache Software License 2.0 across labs, papers, and developer communities, ranked by signal.
- authored by Apache Software Foundation 100%
- instance of Aleph Alpha 95%
- instance of North Mini Code 95%
- used by Granite Embedding Multilingual R2 95%
- instance of Muse Glimmer 90%
- used by Muse Glimmer 90%
- used by DeepInfer: Open-Source Deep Learning Deployment Toolkit for Image-Guided Therapy 90%
- used by tencent/Hy3 90%
- instance of Kimi k3 90%
- instance of tencent/Hy3 90%
- used by Kolibri 90%
- instance of Muse Spark 90%
31 day(s) with sentiment data
-
Tether AI Research releases BotPsy-0.8B agent model for on-device use
Tether AI Research has released BotPsy-0.8B, an open-source, on-device agent model built upon Qwen3.5-0.8B. This new model significantly outperforms its base and other models up to 1.2B parameters on various tool-use an…
-
MBZUAI releases open 29B world model PAN 2 for physical AI · 3 sources tracked
The Institute of Foundation Models (IFM) at MBZUAI has released PAN 2, an open-source 29-billion parameter world model designed for physical AI applications. This model, built on a diffusion transformer architecture, ca…
-
Ricursive's Ripple-65B Preview benchmarked on agentic coding tasks
Ricursive's Ripple-65B Preview model was benchmarked on agentic coding tasks, achieving a score of 0.028 on the Gauntlet sb-7.2 benchmark and scores of 0.0858 and 0.0770 on Forge 2.0. These results are compared against …
-
New Honeypot System Talos Protects Industrial Controllers from AI Agents
Researchers have developed a system called Talos, designed to protect industrial control systems from autonomous language model agents. Talos acts as an inline gate, analyzing agent requests in real-time and redirecting…
-
Developer modifies Qwen3.5-2B model to prevent token-consuming loops
A developer has modified a small 2-billion parameter language model, Qwen3.5-2B, to address its tendency to get stuck in repetitive verification loops, which consume excessive token budgets. The first modification, name…
-
Open-source Canva alternative 'Digital Canvas' launched with agent tools
Panth Patel has developed an open-source alternative to Canva called Digital Canvas, which is free to use and requires no login. The editor stores designs as individual .dcanvas files, allowing users to control where th…
-
KVARK Sparkle 1.1 model enhances document Q&A and reduces hallucinations
KVARK has released Sparkle 1.1, an updated 27B parameter model focused on improving document question-answering capabilities. This new version demonstrates enhanced performance in accurately answering questions based on…
-
Pinky tool grants mobile agents access to Android and iOS devices
A new tool called Pinky, released under the Apache-2.0 license, allows mobile agents to access data from Android (versions 10-17) and iOS (version 17.4+) devices. This access is facilitated through standard input/output…
-
Developer releases 400M parameter model for in-game NPC judgment calls
A developer has created Ettin-400M, a 400 million parameter model designed for making judgment calls in video games. This model, released under the Apache-2.0 license, can process scenes and questions to determine proba…
-
MuJoCo README translated to Chinese to simplify setup
A Chinese translation of the MuJoCo README has been published, aiming to simplify the setup process for Chinese-speaking users. The translated repository includes a quick-start guide that reorders and clarifies the orig…
-
NVIDIA GPU Kubernetes Management Plane NVKE translated into Chinese
A Chinese translation of NVIDIA's NVKE (NVIDIA Kubernetes Engine) management plane has been released, aiming to simplify the setup and operation of Kubernetes clusters on NVIDIA GPU hardware. The translated repository, …
-
Zyphra releases ZONOS3-Mini open-source TTS for real-time local use
Zyphra has released ZONOS3-Mini, an open-source text-to-speech model designed for real-time performance on local devices. This model boasts 100 million active parameters and supports zero-shot voice cloning and word-lev…
-
Mistral AI releases 'Le Chonk' 1T-parameter model for self-hosting
Mistral AI has released "Le Chonk," a 1 trillion parameter, multimodal model with a 1 million token context window, available for download. The model is offered under a modified MIT license that permits commercial use a…
-
Sage-A1.5 model runs on consumer GPUs with careful configuration
The Sage-A1.5 model, developed by Shanghai Artificial Intelligence Laboratory and InternScience, is an open-source 35 billion parameter mixture-of-experts model. It can run on consumer hardware like an RTX 4090 due to i…
-
Nous Research fine-tunes Qwen3.8-27B to slash propaganda by 98%
Nous Research has released Arete-27B, an open-weight model fine-tuned from Alibaba's Qwen3.8-27B. This new model significantly reduces propaganda in responses, dropping it from 22% to 0.9% on Nous's internal evaluation.…
-
Inkling 2 Small: A Developer's Checklist Before Committing to a New Model
Thinking Machines Lab has released Inkling 2 Small, a 594B parameter model with 17B active parameters per token, scoring 31 on the Artificial Analysis Intelligence Index. Currently, access is limited to the Tinker platf…
-
Thinking Machines releases 594B Inkling 2 Small, plans open weights
Thinking Machines has released Inkling 2 Small, a 594 billion parameter model that achieves frontier-level performance on benchmarks. The model utilizes a mixture-of-experts architecture, with only 17 billion parameters…
-
Researchers find limited KV cache reduction potential in standard transformer layers
Researchers investigated the potential for KV cache reduction in standard transformer models, inspired by LoopVQ's success with looped transformers. They found that while some redundancy exists between adjacent layers i…
-
Graphium AI tool standardizes research methods from papers
Graphium is a new open-source desktop application designed to help researchers compare experimental procedures from scientific papers. The tool uses AI to extract and standardize methods, operations, materials, instrume…
-
New transliteration models aim to simplify mixed-script typing
Developers have created UnimeType, a new input method and browser converter, to address the challenges of mixed-script typing. They trained two specialized transliteration models, Transliteration 2.0 (0.8B and 4B), base…