Twitxy Saved

Jonas Geiping

Verified account This account is verified. Learn more

@jonasgeiping

Machine Learning Researcher in Tübingen at the ELLIS Institute & Max Planck for Intelligent Systems // Working on Safety & Efficiency of modern ML
Tübingen AI Center, Germany jonasgeiping.github.io Joined September 2021
849 Following 6.1K Followers 813 Posts

Latest posts 5

Pinned post

Jonas Geiping Verified account @jonasgeiping
We’re training models wrong and it’s due to chatGPT. Even the modern coding agents used daily still use message-based exchanges: They send messages to users, to themselves (CoT) and to tools, and receive messages in turn.

This bottlenecks even very intelligent agents to a single stream. The models cannot read while writing, cannot act while thinking and cannot think while processing information.

In our new paper, see below, we discuss LLMs with parallel streams. We show that multi-stream LLMs can …
🔵Be created by instruction-tuning for the stream format

🔵Simplify user and tool use UX removing many pain points with agents and chat models (such as having to interrupt the model to get a word in)

🔵Multi-Stream LLMs are fast, they can predict+read tokens in all streams in parallel in each forward pass, improving latency

🔵 LLMs with multiple streams have an easier time encoding a separation of concerns, improving security

🔵 LLMs with many internal streams provide a legible form of parallel/cont. reasoning. Even if the main CoT stream is accidentally pressured or too focused on a particular task to voice concerns, other internal streams can subvocalize concerns that would otherwise not be verbalized.

Does this sound related to a recent thinky post :) - Yes, but I don’t feel so bad about being outshipped with such a cool report on their side by 23 hours. I’ll link a 2nd thread below with a more direct comparison. I actually think both are complementary in interesting ways.
45 replies 168 reposts 1.4K likes 165K views View on X
Jonas Geiping Verified account @jonasgeiping
We're looking for new colleagues in Tübingen!
ELLIS Institute Tübingen Verified account @ELLISInst_Tue Oct 1
🚀 The ELLIS Institute Tübingen is hiring Principal Investigators as Hector Endowed Fellows!
Enjoy globally competitive packages, full academic freedom & possible co-appointments w/ @MPI_IS and the Tübingen AI Center (@uni_tue).
📍 Tübingen, 🇩🇪
📅 Apply by Dec 13
🔗 institute-tue.ellis.eu/jobs/PI-call-2…
#AI #MachineLearning #Hiring
View quoted post on X
2 replies 1 reposts 23 likes 1.6K views View on X
Jonas Geiping Verified account @jonasgeiping
This form of animism (see row 3), especially seen in Claude-generation models, is right now the most tiring model-speech quirk to me, with endless documents where "chapters ask", "benchmarks question", "answers matter for the rules" ...

And once more the question, is this just a pattern that Anthropic (or its human preference raters) originally prefered, or is it self-developed?
Thomas Hikaru Clark @thomashikaru Oct 22
We collected matched responses from a range of frontier, near-frontier, and pre-frontier models in chat, agentic coding, and essay-writing settings, ranging from Llama-base to Claude Fable 5 and GPT-5.6 Sol.
View quoted post on X
3 replies 0 reposts 63 likes 3.3K views View on X
Jonas Geiping Verified account @jonasgeiping
This next paper is a bit quirky, but here goes:

Introspection is hard to study in both humans and LLMs. But, we do have internal senses that are verifiable, like proprioception (the sense of where your arm is right now).

What would proprioception look like in language models?

As a digital analogue to the sense of position, a model could verbalize which expert activated in a certain layer for any previous token.

It turns out that LLMs can learn a range of such proprioception skills where they understand and report on their internals. We can calibrate and score these reports, and evaluate whether training for proprioception changes model behaviors.
14 replies 70 reposts 469 likes 23K views View on X
Jonas Geiping Verified account @jonasgeiping
You should find someone who loves you as much as Claude loves Benjamini-Hochberg testing

Purely incidentally, I am sure, the number of submissions to ICLR describing the use of this test in their abstract has quadrupled...

This abstract, in which Claude was tasked to prove AI detection is impossible was especially comical to me