ActionLayer (@ActionLayer) header
ActionLayer (@ActionLayer) profile picture

The First Action Layer for Autonomous Agent 🔗https://teapawt.org/t.co/k8FIVLDV75

San Francisco, CA
Joined April 2024
Photo by ActionLayer (@ActionLayer)Photo by ActionLayer (@ActionLayer)Photo by ActionLayer (@ActionLayer)Photo by ActionLayer (@ActionLayer)Photo by ActionLayer (@ActionLayer)Photo by ActionLayer (@ActionLayer)Photo by ActionLayer (@ActionLayer)Photo by ActionLayer (@ActionLayer)Photo by ActionLayer (@ActionLayer)Photo by ActionLayer (@ActionLayer)
The Sam Altman Interview You know him as the CEO of OpenAI — but he's also an avid writer. We spoke not once but twice about how Sam captures ideas, clarifies his thinking, edits his writing, decides what to work on, and uses ChatGPT. Timestamps: 1:47 Will LLMs change how we write? 8:39 How does Sam use ChatGPT? 11:26 How Sam became less anxious 17:24 Sam once dreamed of being a novelist 18:37 Lessons from Peter Thiel 21:35 Lessons from Paul Graham 26:02 The book Sam Altman wants to write 28:37 Advice for startup founders 30:20 How Y Combinator shapes OpenAI 35:55 How Sam chose to work on AGI 37:35 Writing strategy memos at OpenAI 41:34 Why isn’t ChatGPT a better storyteller? 44:20 Sam's obsessive note-taking method 47:12 Will AI put writers out of work?
43
214
28
1,435
240,073
🎮 We're helping shape the future of gaming through our partnership with @solanalabs! At the center of this collaboration is Gameshift, which provides the full slate of Web3 primitives and actions that games require. Learn more → goo.gle/4gACo0f
41
158
56
458
113,755
How can #robots remember? 🤖 💭 For robots to understand and respond to questions that require complex multi-step reasoning in scenarios over long periods of time, we built ReMEmbR, a retrieval-augmented memory for embodied robots. 👀 Technical deep dive from #NVIDIAResearch ➡️ developer.nvidia.com/blog/us…
14
42
10
171
30,588
ActionLayer retweeted
Transformer by hand✍️ Excel ~ I designed this exercise to show the core math of a Transformer model is to combine columns (attention), combine rows (feed forward), and repeat.👇Join the 'AI Math' community. 👇Download xlsx.
4
210
7
1,065
66,934
Convolution Neural Network Cheat Sheet
11
385
7
2,820
278,640
ActionLayer retweeted
Can LLMs learn to "phone a friend?" 🧵 MIT CSAIL’s new "Co-LLM" algorithm can pair a general-purpose base LLM w/a more specialized model & help them work together. It reviews each token & sees where it needs to call upon an expert, leading to more accurate & efficient replies to medical prompts and math & reasoning problems: bit.ly/4gqx0wS
4
34
6
132
14,286
ActionLayer retweeted
Agents in Software Engineering Provides a comprehensive overview of frameworks of LLM-based agents in software engineering. arxiv.org/abs/2409.09030
7
106
7
467
42,838
ActionLayer retweeted
How does OpenAI train the Strawberry🍓 (o1) model to spend more time thinking? I read the report. The report is mostly about 𝘸𝘩𝘢𝘵 impressive benchmark results they got. But in term of the 𝘩𝘰𝘸, the report only offers one sentence: "Through reinforcement learning, o1 learns to hone its chain of thought and refine the strategies it uses." I did my best to understand this sentence. I drew this animation to share my best understanding with you. The two key phrases in this sentence are: Reinforcement Learning (RL) and Chain of Thought (CoT). Among the contributors listed in the report, two individuals stood out to me: Ilya Sutskever, the inventor of RL with Human Feedback (RLHF). He left OpenAI and just started a new company, Safe Superintelligence. Listing Ilya tells me that RLHF still plays a role in training the Strawberry model. Jason Wei, the author of the famous Chain of Thought paper. He left Google Brain to join OpenAI last year. Listing Jason tells me that CoT is now a big part of RLHF alignment process. Here are the points I hope to get across in my animation: 💡In RLHF+CoT, the CoT tokens are also fed to the reward model to get a score to update the LLM for better alignment, whereas in the traditional RLHF, only the prompt and response are fed to the reward model to align the LLM. 💡At the inference time, the model has learned to always start by generating CoT tokens, which can take up to 30 seconds, before starting to generate the final response. That's how the model is spending more time to think! There are other important technical details missing, like how the reward model was trained, how human preferences for the "thinking process" were elicited...etc. Finally, as a disclaimer, this animation represents my best educated guess. I can't verify the accuracy. I do wish someone from OpenAI can jump out to correct me. Because if they do, we will all learn something useful! 🙌
30
227
17
1,047
89,817
ActionLayer retweeted
Today would have been the 83rd birthday of Dennis Ritchie, inventor of C and co-creator of Unix: zd.net/2creeZi (Image v/@SanderFocus, article v/@jperlow & @ZDNET)
3
95
2
289
14,793
ActionLayer retweeted
Protect your swaps from sandwich attacks. 🛡️ You can now safeguard your swap transactions on Jupiter with one click. By turning on ‘MEV Protect’, your transactions will be sent directly to a @jito_labs validator, reducing the chances of your transactions getting frontrun. Here’s how: 👇
90
182
43
984
116,545
ActionLayer retweeted
LLMs struggle with fine-grained in-line citations in long-context scenarios. This is a really useful capability but current long-context LLMs don't do so well on fine-grained in-line citations. This new work synthesizes a large-scale SFT dataset with off-the-shelf LLMs to improve long-context question answering with citations. It trains 8B and 9B parameter models that enhance citation generation capabilities from lengthy contexts while improving response correctness. Claims to even surpass GPT-4o on their proposed LongBench-Cite benchmark.
4
35
2
164
17,616
ActionLayer retweeted
76-page survey paper on Prompting Techniques ✨ Explores structured understanding and taxonomy of 58 text-only prompting techniques, and 40 techniques for other modalities. 📌 The paper focuses on discrete prefix prompts rather than cloze prompts, because prefix prompts are widely used with modern LLM architectures like decoder-only models. It excludes soft prompts and techniques using gradient-based updates. 📌 The paper identifies 58 text-based prompting techniques broken into 6 major categories: 1) In-Context Learning (ICL) - learning from exemplars/instructions in the prompt 2) Zero-Shot - prompting without exemplars 3) Thought Generation - prompting the LLM to articulate reasoning 4) Decomposition - breaking down complex problems 5) Ensembling - using multiple prompts and aggregating outputs 6) Self-Criticism - having the LLM critique its own outputs 📌 For ICL, it discusses key design decisions like exemplar quantity, ordering, label quality, format, and similarity that critically influence output quality. It also covers ICL techniques like K-Nearest Neighbor exemplar selection. 📌 Extends the taxonomy to multilingual prompts, discussing techniques like translate-first prompting and cross-lingual ICL. It also covers multimodal prompts spanning image, audio, video, segmentation, and 3D modalities. 📌 More complex techniques like agents that access external tools, code generation, and retrieval augmented generation are also taxonomized. Evaluation techniques using LLMs are discussed. 📌 Prompting issues like security (prompt hacking), overconfidence, biases, and ambiguity are highlighted. Two case studies - benchmarking techniques on MMLU and an entrapment detection prompt engineering exercise - are presented.
20
132
12
841
136,665
ActionLayer retweeted
Noisy information is not always bad for RAG systems. Well, it depends on what kind of noise. Finally, there is a benchmark (NoiserBench) to measure how different kinds of noisy information affect RAG's performance. It's interesting that out of the different kinds of beneficial noise (e.g., semantic, datatype, and illegal sentence), illegal sentence noise exhibits the most improved model performance across models and datasets. There are many other interesting results in the paper that you will find useful if you are building RAG systems. "While harmful noise generally impairs performance, beneficial noise may enhance several aspects of model capabilities and overall performance"
4
48
7
209
62,333
ActionLayer retweeted
🚨 REPORT: @Solana surpasses all other chains in monthly NFT blockchain user adoption. - Users: 106K - Dominance: 32%
23
84
8
553
29,400
ActionLayer retweeted
Useful Prompting technique. Simply ask the LLM to re-read the question - significantly boosts LLM reasoning across diverse tasks and model types. 💡 Repeats question input twice in prompt, unlocks latent reasoning potential **Problem** 🤔: Decoder-only LLMs with unidirectional attention struggle with nuanced reasoning tasks due to limited global understanding of input questions. **Key Insights from this Paper 💡**: • Re-reading (RE2) input enhances reasoning by improving question comprehension • Enables "bidirectional" understanding in unidirectional LLMs • Compatible with existing thought-eliciting prompting methods • Effective across various LLM types and reasoning tasks **Solution in this Paper** 🔍: • Introduces RE2 (Re-Reading) prompting method: - Repeats question input twice in prompt - Enhances input understanding before reasoning - Allows tokens to attend to full context in second pass • Compatible with Chain-of-Thought and other prompting techniques • Applicable to zero-shot, few-shot, and self-consistency settings **Results** 📊: • Consistent improvements across 14 datasets and 112 experiments • Effective for both instruction-tuned (ChatGPT) and non-tuned (LLaMA) models • Increases n-gram recall between generation and input question • Most effective when reading question twice
36
161
25
1,037
141,435
ActionLayer retweeted
MemLong utilizes an external retriever for retrieving historical information which enhances the capabilities of long-context LLMs. It consistently outperforms other SoTA LLMs on long-context benchmarks and can extend the context length on a single 3090 GPU from 4k up to 80k.
8
73
1
305
27,329
ActionLayer retweeted
#Otd in 1955, the term “artificial intelligence” was coined in a conference proposal: stanford.io/2WJJJGN
7
161
23
525
49,804
ActionLayer retweeted
PayPal's PYUSD stablecoin recently surpassed a $1 billion market cap! It has increased by 271% since its launch on Solana, with Solana PYUSD accounting for 88% of that growth. Why did PayPal choose to build on Solana and what’s driving PYUSD adoption? 🧵
177
315
87
1,715
390,786
This was a cool listen. I think Cloud+AI is increasingly making the @levelsio -style model of a scrappy solo serial micro-entrepreneur viable, allowing one person to spin up and run a number of companies that generate income, possibly well into billion-dollar valuations.
Here's my conversation with Pieter Levels (@levelsio), self-taught developer and entrepreneur who designed, programmed, shipped, and ran over 40 startups, many of which are hugely successful. In most cases, he did it all by himself, while living the digital nomad life in over 40 countries and over 150 cities. This was a fun and inspiring conversation! It's here on X in full, and is up on YouTube, Spotify, and everywhere else. Links in comment. Timestamps: 0:00 - Introduction 2:03 - Startup philosophy 9:34 - Low points 13:03 - 12 startups in 12 months 19:55 - Traveling and depression 32:34 - Indie hacking 36:37 - Photo AI 1:12:53 - How to learn AI 1:21:30 - Robots 1:29:47 - Hoodmaps 1:53:52 - Learning new programming languages 2:03:24 - Monetize your website 2:09:59 - Fighting SPAM 2:13:33 - Automation 2:24:58 - When to sell startup 2:27:52 - Coding solo 2:33:54 - Ship fast 2:42:38 - Best IDE for programming 2:52:09 - Andrej Karpathy 3:01:34 - Productivity 3:15:21 - Minimalism 3:24:07 - Emails 3:31:20 - Coffee 3:39:05 - E/acc 3:41:21 - Advice for young people
155
519
75
5,635
963,843
ActionLayer retweeted
Eco growth is accelerating as founders who started in @Solana hackathons are now giving back as sponsors: @0xdirichlet-@sphere_labs @crispheaney-@DriftProtocol @aliquotchris-@leapxyz As well as new Colosseum portfolio startups: @metaproph3t-@MetaDAOProject Study compounding
8
17
3
97
17,433