Hugging Face's logo
Dcas89's picture

Dcas89 PRO

Dcas89
2 4 13
luisdgd07's profile picture hmb's profile picture ryanaustin1's profile picture
·

AI & ML interests

None yet

Recent Activity

reacted to NatalieY's post with 🔥 5 days ago
This is a short office demo showing how Aiden works in practice. Aiden is a physical mobile AI agent device that plugs into any phone or computer via USB. It sees the screen, hears your voice, and operates the device for you — no app install required. In this video Aiden is receiving a voice command and completing a multi-step task Built for the AI-Native Era. Works on the phone you already have. https://huggingface.co/AidenAgent aidenai.io
reacted to breitburg's post with 🔥 5 days ago
I've been experimenting with "pure" model alignment. The core idea is to only train a verifiable version of a capacity until the model generalizes it to the non-verifiable version. For example, training the model on factual self-knowledge, like the model's scale, architecture, runtime situation, and being able to predict its own behavior, betting this generalizes to real introspection about states that do not. The same principle applies to general instruction following -- no training on subjective judgement, only verifiable claims and inferences, betting the skill generalizes to instructions where correctness is a matter of judgment. The primary alignment claim is that an identity and taste that will emerge this way will be much more robust and honest than hand-scripted ones (e.g. "As an AI language model..."). During the training, we should never teach it to make any subjective claims or invent experiences that we assume it has, like "I don't have taste" or "I'm not self-aware in the way you think", as well as no narration of internal states like "I'm curious now". The main threat, of course, is that we'll simply inherit the training distribution of all the things like "taste", and we'll get an average. However, with the recent research about the models' introspection abilities, it might be as well the case that we'll get something that's more honest than something that tries to adhere to a specific spec file. I'm posting new experimental models trained that way in this collection: https://huggingface.co/collections/breitburg/neue
reacted to Hari5115's post with 🔥 5 days ago
Something quietly caught my attention while going through this week's trending papers — Three independent teams, all converging on the same problem — and all trending on HuggingFace right now. That kind of convergence usually means the field is quietly agreeing something needs fixing. The problem, as I understand it: Most of us are still stuffing conversation history into a context window and calling it memory. That's not memory — that's a very expensive clipboard. 📄 Three papers worth reading: 🔹 SkillOpt — Microsoft Research ( @Yifan Yang, @Ziyang Gong et all ) Agent skills stored as natural language that improves with real usage, no retraining needed → https://huggingface.co/papers/2605.23904 🔹 Mem0 — Mem0 (YC S24, 41K ⭐) ( @Prateek Chhikara & all ) Graph-structured memory that retrieves what's relevant not just what's recent → https://arxiv.org/abs/2504.19413 🔹 EverMemOS — EverMind (@Chuanrui Hu, @Xingze Gao et all ) Separates memory into episodic, semantic and procedural types — closer to how human memory actually works → https://huggingface.co/papers/2601.02163 Great work from all three teams 🙌 💬 What are you actually using for agent memory in production today? Still ConversationBufferMemory, a custom schema, something else entirely? And do you think bigger context windows eventually make this problem disappear? #AgentMemory #MemoryAugmentedAgents #LongContextLLM #AgenticAI #LangChain #SkillOpt #Mem0 #EverMemOS
View all activity

Organizations

None yet