Nibo.AI (@nibo_ai) profile picture

spawning LLM agents for redundant tasks

Baden, Switzerland
Joined October 2017
Photo by Nibo.AI (@nibo_ai)Photo by Nibo.AI (@nibo_ai)Photo by Nibo.AI (@nibo_ai)Photo by Nibo.AI (@nibo_ai)Photo by Nibo.AI (@nibo_ai)Photo by Nibo.AI (@nibo_ai)
⚔️ Ai
Are LLMs good at handling geo-political conflicts? Nice deep-dive on this topic in @MLamparth's paper: Human vs. Machine: Language Models and Wargames simulating a US-China conflict TL;DR: Please don’t do this 😬
1
23
Nibo.AI retweeted
Wow! This is cool, for anyone using Self-consistency-like methods. This paper discovers a non-monotonic relationship between the number of LLM calls and performance. Contrary to what one might intuitively expect, making more calls to a LLM not always = performance improvements
This surprised us. If you call #chatgpt multiple times and take the consensus, the quality of the answer can get worse as the # of #LLM calls increases. We explain why this happens in our new paper and also show how to estimate the optimal # of LLM calls arxiv.org/pdf/2403.02419.pdf
2
5
11
4,977
😮‍💨🧘
Enough detour, finally reaching the culmination of funny zero-shot CoT keywords, for PaLM 2 at least: Take a deep breath and work on this problem step-by-step
95
Nibo.AI retweeted
The big bonus, it can improve CoT
1
1
1
26
Early days of task breakdown, I was using Plan and Solve
Plan and Solve it probably the most powerful improvement over Zero-Shot CoT's "Let's think step by step". Actually so powerful that sometimes it matches few-shot CoT
15
Plan and Solve it probably the most powerful improvement over Zero-Shot CoT's "Let's think step by step". Actually so powerful that sometimes it matches few-shot CoT
1
3
3
9
3,456
STEP-BACK PROMPTING is a method to improve the reasoning abilities of LLMs by prompting to first abstract high-level concepts / principles before reasoning towards the solution. Demonstrated substantial performance++ across a range of challenging reasoning-intensive tasks.
2
1
1
2
344
A funny one from @SFResearch : The FlipFlop effect Introduced the FlipFlop experiment* : a multi-turn conversation setup where LLMs first provide an answer to a classification task and are then challenged with a follow-up phrase like "Are you sure?". authors here: @PhilippeLaban @CaimingXiong , @murakhovska, @jasonwu0731
1
1
5
172
I was using GPT4 Advanced Data Analysis to explore some data and it suggested to do Sentiment Analysis. It worked from the 1st try (!) but the moment I saw the results it was obviously a cool situation where the "pure" LLM would do a much more satisfying job. Some caveats bellow:
3
5
12
1,033
2. I found that when using the above prompt, GPT4 provided better outputs compared to Claude. However, I still preferred using Claude due to its 100k context window. This allowed me to quickly classify a few hundred rows without the need to divide them into multiple prompts.
1
1
5
136
So when using llm agents, the ability to know when they need help is crucial for human-AI collaboration. We got inspired by a cool idea from a Deepmind robotics paper
2
16
4
30
9,582
beta testing...
So when using llm agents, the ability to know when they need help is crucial for human-AI collaboration. We got inspired by a cool idea from a Deepmind robotics paper
29
Enriching inter-departmental operations in the modern workplace linkedin.com/in/michael-eder… #automation #worktogether #kanban #collaboration
2
What's the most common UI you interact with every day? Design insights from @ioanahalunga - Transitioning from a classic interface to a #conversationalUI. blog.nibo.ai/chat-inputs-vs-… #chatbot #Bot #Automation #ux #uidesign #design
1
1
A friendly #bot 🤖, a four in one matching #algorithm 🐙 and a #cloud #automation engine ⚡ walk into a bar... buff.ly/2D9uLiK