Knut Jägersberg (@jagersbergknut) 's Twitter Profile
Knut Jägersberg

@jagersbergknut

Content Strategy & AI

@[email protected]

huggingface.co/KnutJaegersberg

ID: 1010498049058201600

linkhttps://www.linkedin.com/in/knut-jägersberg calendar_today23-06-2018 12:21:23

91,91K Tweet

5,5K Takipçi

4,4K Takip Edilen

Daniel Jeffries (@dan_jeffries1) 's Twitter Profile Photo

Good idea for both post-o1 next-gen thinking systems and agentic memory systems and coincidentally one of the approaches we are looking at strongly at the moment. arxiv.org/abs/2409.10038…

nat://TheAIObserverX (@theaiobserverx) 's Twitter Profile Photo

Researchers introduce a new evaluation framework for testing affective cognition in AI. By generating 1,280 scenarios based on psychological theory, they test GPT-4, Claude-3, Gemini-1.5-Pro, and 567 humans. Findings reveal models often align with human intuitions, sometimes even

Justin Chih-Yao Chen (@cyjustinchen) 's Twitter Profile Photo

Aggregation & refinement improve LLM reasoning, but aggregation saturates, while refinement has 3 issues: 1) over-correction for easy problems 2) fails to localize+fix its own errors 3) insufficient number of refinement iteration for hard problems 🚨Multi-Agent, Iterative,

Aggregation & refinement improve LLM reasoning, but aggregation saturates, while refinement has 3 issues:

1) over-correction for easy problems
2) fails to localize+fix its own errors
3) insufficient number of refinement iteration for hard problems

🚨Multi-Agent, Iterative,
Bill Yuchen Lin 🤖 (@billyuchenlin) 's Twitter Profile Photo

Congratulations to Qwen 👏 Their Qwen2.5-72B model has become the top-performing open model on ZeroEval, which evaluates performance across multiple challenging reasoning tasks.​​​​​​​​​​​​​​​​ 🔗 on Hugging Face: hf.co/spaces/allenai…

Congratulations to <a href="/Alibaba_Qwen/">Qwen</a> 👏
Their Qwen2.5-72B model has become the top-performing open model on ZeroEval, which evaluates performance across multiple challenging reasoning tasks.​​​​​​​​​​​​​​​​ 

🔗 on <a href="/huggingface/">Hugging Face</a>: hf.co/spaces/allenai…
Wolfram Ravenwolf 🐺🐦‍⬛ (@wolframrvnwlf) 's Twitter Profile Photo

As a German (and thus European) AI Engineer, consultant, and researcher, I fully agree with and support this open letter. The EU needs AI! It is the key technology that will determine whether Europe remains relevant and prosperous in the future. Regulatory certainty is necessary.

Migel Tissera (@migtissera) 's Twitter Profile Photo

I just created a single-turn dataset of 250K samples using DeepSeek-V2.5 API. Total cost? $200. Unreal. Intelligence is really approaching the cost of electricity yo!

Delip Rao e/σ (@deliprao) 's Twitter Profile Photo

OpenAI is what I call a “parasitic science organization”. They take stuff from the open science community, use them opaquely, and profit from it, without giving much back to open science. And if you point out, you get gaslit with plausible deniabilities. We all remember the

Ross Taylor (@rosstaylor90) 's Twitter Profile Photo

I don’t subscribe to the view that “ideas are cheap, execution is everything”, in the sense that good intuition is hard to come by. But what is true is that new ideas are incredibly fragile. At an organisation that prioritises reactive plays, these ideas will die without

Noam Brown (@polynoamial) 's Twitter Profile Photo

.OpenAI is hiring ML engineers for a new multi-agent research team! We view multi-agent as a path to even better AI reasoning. Prior multi-agent experience isn't needed. If you'd like to research this area with Kevin Stone and me fill out this form: jobs.ashbyhq.com/openai/form/oa…