Surge AI (@HelloSurgeAI) / X

Surge AI

679 posts

Surge AI

@HelloSurgeAI

Our mission is to raise AGI with the richness of humanity — curious, witty, imaginative, and full of breathtaking brilliance.

Joined June 2020

Surge AI
@HelloSurgeAI
Nov 11, 2025
Everyone's acting like models are ready to replace humans in work settings. We put that to the test by creating an entire company and having 9 models act as a customer service agent handling 150 tickets and requests of increasing complexity. Verdict: without common sense,
RL Environments and the Hierarchy of Agentic Capabilities
From surgehq.ai
247K
Surge AI
@HelloSurgeAI
Mar 27, 2023
RLHF helps to build state-of-the-art models like ChatGPT. Did you know that training RLHF LLMs involve 4 key steps? Here’s an illustrated guide to the process: 1 of 5
121K
Surge AI
@HelloSurgeAI
Apr 20, 2023
Awesome RLHF Nice collection of research papers for RLHF, including code links, datasets, blogs, etc. github.com/opendilab/awes…
27K
Surge AI
@HelloSurgeAI
Apr 18, 2023
Open-source RLHF implementations are on the rise! DeepSpeed Chat and ColossalChat are two open-source RLHF pipeline implementations announced in just the past couple of weeks. Here’s why they matter:
59K
Surge AI
@HelloSurgeAI
Apr 21, 2023
Every week we cover key papers in RLHF LLMs. Last week we covered InstructGPT, and it got a lot of interest. We continue this week with DeepMind’s GopherCite paper. Here’s what you need to know in 5 tweets:
50K
Surge AI
@HelloSurgeAI
May 9, 2023
Human feedback is important to train safe and helpful LLMs. Did you know there is now a large taxonomy of methods that leverage human feedback? This recent paper provides a comprehensive overview of recent methods. arxiv.org/abs/2305.00955
32K
Surge AI
@HelloSurgeAI
Apr 28, 2023
Every week we cover key papers in RLHF and LLMs. Today’s paper explores whether humans + LLMs working together can outperform either alone on difficult tasks. Here’s the paper summary in 5 tweets:
37K
Surge AI
@HelloSurgeAI
May 5, 2023
Every week we cover key papers in RLHF and LLMs. Today’s paper explores how RLHF can help train helpful and harmless assistants. Here’s the paper summary in 5 tweets:
33K
Surge AI
@HelloSurgeAI
May 25, 2023
State of GPT and RLHF LLMs Great talk by @karpathy on the state of LLMs and the RLHF training pipeline: build.microsoft.com/en-US/sessions… Here are a few additional readings to learn more about RLHF LLMs:
37K
Surge AI
@HelloSurgeAI
May 26, 2023
Every week we cover key papers in RLHF and LLMs. Today’s paper explores how instruction finetuning can help improve the performance and usability of pretrained language models. Here’s the summary in 5 tweets:
25K
Surge AI
@HelloSurgeAI
May 19, 2023
Every week we cover key papers in RLHF and LLMs. Today we explore WebGPT - browser-assisted question-answering with human feedback. Here’s the paper summary in 5 tweets:
31K
Surge AI
@HelloSurgeAI
Mar 24, 2023
What are the benefits of training LLMs with RLHF? The best way to show this is with examples. Let’s have a look:
55K
Surge AI
@HelloSurgeAI
Apr 14, 2023
Last week, we covered key papers in RLHF LLMs. It got a lot of interest, so we will do a few paper explainers. This time we discuss InstructGPT:
30K
Surge AI
@HelloSurgeAI
Apr 7, 2023
Brief History of RLHF LLMs Here are 5 important works to help you learn about RLHF LLMs:
39K