Most AI is not truly open-source, even if they say they are "open", because they do not let you change the original training data, which is secret.

#Tülu and #OLMo are AI models more open than most, with available for download and modification: model weights, code, and (for OLMo) much of the training data.

They offer free AI chat, which for my coding did well, but for asking about a spiritual author it hallucinated badly, so like all #AI, is not perfect.

https://playground.allenai.org/

#AllenAI

During my intro at #Hack4Socialgood this morning, I used 🤗 Hugging Face playground to quickly show #Apertus prediction for the hackathon compared to the Olmo model from #AllenAI and #EuroLLM

Prompt and response in the ALT. Try it yourself at https://huggingface.co/playground?modelId=swiss-ai/Apertus-70B-Instruct-2509

merve (@mervenoyann)

Allen Institute for AI(AI2)가 웹 브라우저 사용과 클릭 작업을 수행하는 MolmoWeb을 공개했다. 4B와 8B 모델 및 관련 데이터셋을 함께 제공하며, Apache 2.0 라이선스로 배포해 오픈소스 웹 에이전트 연구와 개발에 중요한 진전이다.

https://x.com/mervenoyann/status/2036474373654118777

#allenai #molmoweb #opensource #webagent #llm

merve (@mervenoyann) on X

AI2 @allen_ai released MolmoWeb: 4B and 8B browser use (clicking) models as well as their datasets 🔥 all under Apache 2.0 license as always 💗

X (formerly Twitter)

merve (@mervenoyann)

Allen Institute for AI(allenai)가 공개한 MolmoPoint-8B 및 MolmoPoint-GUI-8B의 웹 데모가 Hugging Face Spaces에 올라와 체험 가능하다는 알림입니다. 제공된 링크를 통해 MolmoPoint 모델의 동작과 GUI 기반 인터페이스를 직접 확인하고 실험해볼 수 있습니다.

https://x.com/mervenoyann/status/2034346787243172120

#huggingface #allenai #molmopoint #demo #8b

merve (@mervenoyann) on X

they also have demos 🤩 https://t.co/z6mh8rvjJ1 https://t.co/t7f8efLTEp 🙌🏼

X (formerly Twitter)
Olmo Hybrid offers the same performance as Olmo 3 but with 49% fewer training tokens by combining transformers and linear recurrent layers, may offer better expressivity and scalability than pure models.
Released by @allenai with open weights, checkpoints, and code for community use and research #allenai #olmohybrid
https://allenai.org/blog/olmohybrid
Introducing Olmo Hybrid: Combining transformers and linear RNNs for superior scaling | Ai2

Ai2, a non-profit research institute founded by Paul Allen, is committed to breakthrough AI to solve the world’s biggest problems.

AIトレンド速報|最新ニュース & 活用術 (@AI_Bridge_Japan)

Allen AI가 개발한 다국어 OCR 벤치마크 'OlmOCR-Bench'가 Hugging Face의 공식 벤치마크 데이터세트로 공개되었습니다. 이로써 다양한 OCR 모델의 성능을 표준화된 환경에서 비교·평가할 수 있게 되어, 다국어 문자 인식 연구에 중요한 기준을 제시합니다.

https://x.com/AI_Bridge_Japan/status/2026120760423764157

#ocr #benchmark #allenai #huggingface #multilingualai

AIトレンド速報|最新ニュース & 活用術 (@AI_Bridge_Japan) on X

Allen AI (@allen_ai) が開発した多言語OCRベンチマーク「OlmOCR-Bench」が、Hugging Face (@huggingface) の公式ベンチマークデータセットとして公開されました。これにより、様々なOCRモデルの性能を標準化された環境で評価できます。(via @mervenoyann) https://t.co/me42EnGvyN

X (formerly Twitter)

AllenAI công bố bộ mô hình mã nguồn mở “Open Coding Agents”, hỗ trợ lập trình tự động và tương tác đa dạng. Các mô hình đã được chia sẻ trên HuggingFace và kèm bài báo chi tiết. #AI #MachineLearning #OpenSource #AllenAI #MãNguồnMở #CôngNghệ #LậpTrình #NghiênCứuAI

https://www.reddit.com/r/LocalLLaMA/comments/1qoocgn/allenai_released_new_open_coding_models/

OLMo 3: 학습 데이터부터 추론 과정까지 완전히 열린 AI 모델

Allen AI가 학습 데이터부터 추론 과정까지 전체 개발 파이프라인을 공개한 OLMo 3 발표. 완전 오픈 중 최고 성능의 32B 추론 모델과 9.3조 토큰 데이터셋을 소개합니다.

https://aisparkup.com/posts/6760

This is another bunch of cool #allenai open models:

https://allenai.org/blog/olmoearth-models

#ai #genai

OlmoEarth: A new state-of-the-art Earth observation foundation model family | Ai2

OlmoEarth is a new family of industry-leading open foundation models built to make Earth AI practical, scalable, and performant for real-world applications.

LLM -Ai2 SciArena: OpenAI's o3 tops new AI league table for answering scientific questions
https://www.nature.com/articles/d41586-025-02177-7
nonpaywalled: https://archive.fo/0IDls

Ai2 SciArena: An Open Evaluation Platform for Foundation Models in Scientific Literature Tasks
https://arxiv.org/abs/2507.01001
https://allenai.org/blog/sciarena

https://sciarena.allen.ai/

#LLM #Ai2 #SciArena #Chato3 #AllenAI #OpenAI

OpenAI’s o3 tops new AI league table for answering scientific questions

SciArena uses votes by researchers to evaluate large language models’ responses on technical topics.