Jonas Becker

Researcher at @GippLab (@uniGoettingen) and @polizei_nrw_lka who loves #NLP, #AI and food. I also enjoy drinking tea while coding!

[Jonas Becker @BeckerNLP]

🚀 Excited to share our EMNLP 2025 Demo Paper!

MALLM: Multi-Agent Large Language Models Framework

Conduct your experiments on agents discussions, and decisions.

🎥 https://t.co/mgBKg8UlQx

👋 See our poster in Session 11 (Thu, Nov 6 · 16:30–18:00, Hall C)

#EMNLP2025

[Jonas Becker @BeckerNLP]

🚀 Presented at #EACL2026: "Stay Focused: Problem Drift in Multi-Agent Debate"

We identify and mitigate performance drift in ongoing multi-agent interactions.

Key insight: longer debates ≠ better answers - staying focused matters more.

📜 Paper: https://t.co/KmNqCj0AkT

[Jonas Becker @BeckerNLP]

We will present our work "Stay Focused: Problem Drift in Multi-Agent Debate" at #EACL2026 this week.

🗓️ Our session is Friday (Poster Session 6, 11:00-12:30).

If you are around, feel free to stop by so we can have a chat or just say hi.

Paper: https://t.co/KmNqCj0AkT

[Jonas Becker @BeckerNLP]

🚀DimStance: Multilingual Dimensional Stance Analysis

🧠 Stance ≠ just Favor / Against
📏 Models stance on valence (low → high) & arousal (calm → active)

📢English, German, Chinese, Nigerian Pidgin, and Swahili
🌍 Politics, Environment

https://t.co/1XipLoYfjQ

#SemEval2026

[Jonas Becker @BeckerNLP]

🚀 Stay Focused: Problem Drift in Multi-Agent Debate
✅ Accepted at #EACL2026

🗣️ MAD drifts from the original problem over turns
🔎 Analysis across reasoning, knowledge & instruction-following
🛠️ DRIFTJudge (detect) & DRIFTPolicy (mitigate)

📄 Preprint: https://t.co/q3Jwe01Scn

[Jonas Becker @BeckerNLP]

🚀 MALLM: a plug-and-play framework for multi-agent debate.
✅ Accepted as #EMNLP2025 Demo

🔧 144+ configurations out of the box
🔎 Find the best multi-agent setup for your research
📄 Preprint: https://t.co/DrPFdYoUZ6
🧪 Demo: https://t.co/g3Qru5uDLj

[GippLab@Uni-Göttingen @GippLab]

GippLab attending #ACL2025NLP in Vienna this week!

📄 We presented three papers 🙌

🔗 Find the titles and links to all papers in the comments below👇

#ACL #ACL2025 #ACL25

[Jan Philip Wahle @jpwahle]

What a nice birthday gift: ACL Best Resource Paper Award and SemEval Best Task Award. Thanks to all the collaborators who made this possible!

[Jonas Becker @BeckerNLP]

More examples and experiments are available in our new preprint "Stay Focused: Problem Drift in Multi-Agent Debate"

https://t.co/7duaWLDLIN

[Jonas Becker @BeckerNLP]

❓ What is Problem Drift in multi-agent debate?

This example shows how agents start with a good solution. However, it gets worse with longer debates.
One agent induces a logical error in the debate. The other agents agree without skepticism, leading to the wrong solution.

[Jonas Becker @BeckerNLP]

Stay Focused: Problem Drift in Multi-Agent Debate

Multi-agent LLMs are prone to making errors during longer interactions.

Check out how we define this "problem drift", investigate its reasons, and test detection and mitigation strategies at test-time.

https://t.co/q3Jwe02q1V

[Jonas Becker @BeckerNLP]

The ethical alignment of Multi-Agent LLMs (MALLM) collapses during ongoing discussions.

This raises concerns about AI safety, highlighting how multi-agent settings come with novel safety challenges that weren't relevant in single-agent scenarios.

Paper: https://t.co/lh8Ur7Ldgz

[Jonas Becker @BeckerNLP]

🚀MALLM🚀 - Conduct your own multi-agent research by using our new framework for problem-solving:
https://t.co/Pi3wjn8fsP

MALLM comes with a dataset loader, easy yet configurable discussion formats, and an integrated evaluation pipeline. 🥳

[Jonas Becker @BeckerNLP]

Multi-Agent LLMs for Conversational Task-Solving 💬

Contributions:
1) Taxonomy of multi-agent systems for task-solving
2) Multi-agent framework for your studies
3) Identifies three problems with multi-agent systems: Performance, Alignment, Monopolization

https://t.co/nz4a9JUbTV

[Jonas Becker @BeckerNLP]

@OpenAI I am happy to wait a few seconds more if the answer is better. That's a good direction to go.

[Jonas Becker @BeckerNLP]

@yang3kc Nice work! The more information you give in the prompt and the more the model has to care for during the generation, the less focus can be on the actual task. So it makes total sense that reasoning gets worse here.
Would be interesting to see other tasks evaluated like this too

[Jonas Becker @BeckerNLP]

@MatthewBerman It's crazy how easy that is. But even when fixing this, people will discover new ways.

[Jonas Becker @BeckerNLP]

@mckaywrigley I already use @cursor_ai a lot. Within minutes, I got bar charts, line graphs, and correlation matrices for my research project.
I just needed to adjust some little things to make it look nice.

[Jonas Becker @BeckerNLP]

@Megatron_ron That's no breaking news. It's broken news.

[Jonas Becker @BeckerNLP]

8/8 🚀 Challenges:
We identify 9 overarching challenges in text generation.
These are bias, misuse, reasoning, hallucinations, privacy, transparency, interpretability, datasets, and computing.
For each, we survey state-of-the-art research and provide research directions.

[Jonas Becker @BeckerNLP]

📑 Text Generation: A Systematic Literature Review of Tasks, Evaluation, and Challenges

Explore recent advances in text generation since 2017, focusing on five core sub-tasks and highlighting key research gaps.

🔗 Read the paper: https://t.co/97IMuVnS7O

🧵 1/8

[Jonas Becker @BeckerNLP]

7/8 📊 Evaluation:
Researchers heavily rely on automated metrics.
We find that most works use n-gram overlap metrics like BLEU, ROUGE, and METEOR.
We raise awareness about other metrics to complement evaluation (statistical, graph-based, model-based).