Hacker News headlines

2

CodePlan: Repository-level Coding using LLMs and Planning (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

2

Optimal tic-tac-toe with GPT-4 (openai.com)

2 years ago | og_kalu | openai.com | newest

8

Parallelizing non-linear sequential models over the sequence length (arxiv.org)

2 years ago | og_kalu | arxiv.org | best

4

Dalle-3/GPT Screen Capture: Sky Dachshund [video] (youtube.com)

2 years ago | og_kalu | youtube.com | frontpage

1

GPT-3.5-turbo-instruct vs. Stockfish level 5 (reddit.com)

2 years ago | og_kalu | reddit.com | newest

3

Comprehension of Sentences by Bottlenosed Dolphins (sciencedirect.com)

2 years ago | og_kalu | sciencedirect.com | newest

76

Large Language Models for Compiler Optimization (arxiv.org)

2 years ago | og_kalu | arxiv.org | best

2

The Shapes of Stories with ChatGPT (superbowl.substack.com)

2 years ago | og_kalu | substack.com | newest

2

Lingo-1: Exploring Natural Language for Autonomous Driving – Wayve (wayve.ai)

2 years ago | og_kalu | wayve.ai | newest

1

Large Language Model for Science: A Study on P vs. NP (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

1

Large Language Models Are State-of-the-Art Evaluators of Translation Quality (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

1

Unveiling Theory of Mind in LLMs: Parallels to Single Neurons in the Human Brain (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

1

When Less Is More: Investigating Data Pruning for Pretraining LLMs at Scale (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

2

Using GPT-4 to Analyze Medical Records of Patients with Delayed Diagnosis (nih.gov)

2 years ago | og_kalu | nih.gov | newest

1

Open Interpreter – open-source implementation of OpenAI's Code Interpreter (github.com/killianlucas)

2 years ago | og_kalu | github.com | newest

46

Large Language Models as Optimizers. +50% on Big Bench Hard (arxiv.org)

2 years ago | og_kalu | arxiv.org | best

1

Are ChatGPT and GPT-4 Good Poker Players? Yes but Not Game Theory Optimal (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

2

GPT-4 vision finally rolling out to Be my Eyes users (twitter.com/j_stonemountain)

2 years ago | og_kalu | twitter.com | newest

42

Can programming languages boost each other via instruction tuning? (arxiv.org)

2 years ago | og_kalu | arxiv.org | best

2

Do Sequels Often Outgross Their Predecessors? A 23 Year Analysis of 438 Sequels (github.com/ogkalu2)

2 years ago | og_kalu | github.com | newest

1

Box Office Sequel Data/Analysis of the past 23 years (github.com/ogkalu2)

2 years ago | og_kalu | github.com | newest

18

Expanding Transformer size without losing function or starting from scratch (arxiv.org)

2 years ago | og_kalu | arxiv.org | best

1

Link-Context Learning for Multimodal LLMs (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

4

Tool Documentation Enables Zero-Shot Tool-Usage with Large Language Models (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

13

Multimodal Neurons in Pretrained Text-Only Transformers (huggingface.co)

2 years ago | og_kalu | huggingface.co | frontpage

2

From Sparse to Soft Mixtures of Experts. Outperforms Dense/Sparse models (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

1

Skills-in-Context Prompting: Unlocking Compositionality in Large Language Models (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

1

Communicative LLM Agents for Software Development (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

5

GPT-4 Vision (imgur.com)

2 years ago | og_kalu | imgur.com | newest

2

Generating songs with coherent speech and sound effects (suno-ai.notion.site)

2 years ago | og_kalu | notion.site | newest

1

Does Visual Pretraining Help End-to-End Reasoning? (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

1

One Embedder, Any Task: Instruction-Finetuned Text Embeddings (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

40

Model card and evaluations for Claude models [pdf] (anthropic.com)

2 years ago | og_kalu | anthropic.com | best

2

Large Language Models can complete complex non linguistic patterns in context (huggingface.co)

2 years ago | og_kalu | huggingface.co | newest

1

Large Language Models as General Pattern Machines (general-pattern-machines.github.io)

2 years ago | og_kalu | github.io | newest

1

Teaching Arithmetic to Small Transformers (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

5

GPT-4 solves Mystery-o-Matic's Mystery Puzzle of the day (reddit.com)

2 years ago | og_kalu | reddit.com | newest

1

XTrimoPGLM: Unified 100B-Scale Transformer for Deciphering the Protein Language (biorxiv.org)

2 years ago | og_kalu | biorxiv.org | newest

2

Instruct tuned Mixture of Experts LLMs significantly surpass dense counterparts (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

1

Building Cooperative Embodied Agents Modularly with Large Language Models (umass.edu)

2 years ago | og_kalu | umass.edu | newest

1

KokoMind: Can LLMs Understand Social Interactions? (github.com/chats-lab)

2 years ago | og_kalu | github.com | newest

1

KokoMind: Can LLMs Understand Social Interactions? (chats-lab.github.io)

2 years ago | og_kalu | github.io | newest

3

LongNet: Scaling Transformers to 1B Tokens (arxiv.org)

2 years ago | og_kalu | arxiv.org | frontpage

97

With plugins, GPT-4 posts GitHub issue without being instructed to (openai.com)

2 years ago | og_kalu | openai.com | best

2

Demystifying GPT Self-Repair for Code Generation (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

9

Classifier Free Guidance works on LLMs with a significant boost in performance (arxiv.org)

2 years ago | og_kalu | arxiv.org | frontpage

2

Protein-Protein Interaction Prediction Is Achievable with Large Language Models (biorxiv.org)

2 years ago | og_kalu | biorxiv.org | frontpage

2

GPT4GEO: How a Language Model Sees the World's Geography (arxiv.org)

2 years ago | og_kalu | arxiv.org | frontpage

3

On giving AI eyes and ears (oneusefulthing.org)

2 years ago | og_kalu | oneusefulthing.org | newest

2

Kosmos-2: Grounding Multimodal LMMs to the World, demo and model released (arxiv.org)

2 years ago | og_kalu | arxiv.org | frontpage

11

HyenaDNA: Long-Range Genomic Sequence Modeling (context length of 1M tokens) (arxiv.org)

2 years ago | og_kalu | arxiv.org | frontpage

1

Supervised Pretraining Can Learn In-Context Reinforcement Learning (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

11

DeepMind's new Gemini AI will combine LLMs with techniques from AlphaGo (wired.com)

2 years ago | og_kalu | wired.com | frontpage

1

Scaling MLPs: A Tale of Inductive Bias (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

2

Designing Stable and Transferable Sparse Expert Models. First SOTA Sparse LLM (arxiv.org)

2 years ago | og_kalu | arxiv.org | frontpage

1

Inflection debuts new LLM, Outperforms GPT-3.5, Palm-540B on academic benchmarks (inflection.ai)

2 years ago | og_kalu | inflection.ai | newest

1

AudioPaLM: A Large Language Model That Can Speak and Listen (google-research.github.io)

2 years ago | og_kalu | github.io | newest

5

Bing Chat gets GPT-4 Image Input update to select users. Breaks Captcha (twitter.com/sayashk)

2 years ago | og_kalu | twitter.com | newest

1

Do LLMs Understand User Preferences? Evaluating LLMs on User Rating Prediction (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

2

MotionGPT: Finetuned LLMs Are General-Purpose Motion Generators (twitter.com/_akhaliq)

2 years ago | og_kalu | twitter.com | newest

1

LLM with chemistry tools synthesizes catalysts, novel dye, and insect repellent (twitter.com/andrewwhite01)

2 years ago | og_kalu | twitter.com | newest

1

TidyBot: Personalized Robot Assistance with Large Language Models (princeton.edu)

2 years ago | og_kalu | princeton.edu | newest

1

Accuracy of GPT-4 in a Complex Medical Diagnostic Challenge (jamanetwork.com)

2 years ago | og_kalu | jamanetwork.com | newest

7

GPT-4 frustrated with failing browsing tool, searches how to switch the browser (imgur.com)

2 years ago | og_kalu | imgur.com | newest

1

Evidence of Meaning in Large Language Models Trained on Programs (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

223

MusicGen: Simple and controllable music generation (honu.io)

2 years ago | og_kalu | honu.io | best

1

Experimental results from applying GPT-4 to an unpublished formal language (twitter.com/gregorvscheidt)

2 years ago | og_kalu | twitter.com | newest

31

StyleDrop: Text-to-Image Generation in Any Style (styledrop.github.io)

2 years ago | og_kalu | github.io | best

1

LLM Itself Can Read and Generate CXR Images (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

2

Fine-Tuning Language Models with Just Forward Passes (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

71

A Mechanistic Interpretability Analysis of Grokking (alignmentforum.org)

2 years ago | og_kalu | alignmentforum.org | best

2

Tiny Transformer trained for addition and examined with bizarre results (twitter.com/robertskmiles)

2 years ago | og_kalu | twitter.com | newest

2

Improving Factuality and Reasoning in Language Models Through Multiagent Debate (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

3

RecurrentGPT: Interactive Generation of (Arbitrarily) Long Text (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

2

Bing vs Cleverbot (reddit.com)

2 years ago | og_kalu | reddit.com | newest

1

Evidence of Meaning in Language Models Trained on Programs (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

4

Tree of Thoughts: Deliberate Problem Solving with Large Language Models (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

2

SoundStorm: Efficient Parallel Audio Generation. 30s dialogue generated in 2s (google-research.github.io)

2 years ago | og_kalu | github.io | newest

2

Palm 2 – a 340B model trained on 3.6 Trillion tokens of text (cnbc.com)

2 years ago | og_kalu | cnbc.com | newest

1

Symbol tuning improves in-context learning in language models (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

5

TinyStories: How Small Can Language Models Be and Still Speak Coherent English? (huggingface.co)

2 years ago | og_kalu | huggingface.co | newest

2

Code trained LLMs reason better, on benchmarks that have nothing to do with code (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

4

TidyBot: Personalized Robot Assistance with LLMs (princeton.edu)

2 years ago | og_kalu | princeton.edu | newest

2

Microsoft launches Bing chat features incl multimodality, plug-ins, image search (venturebeat.com)

2 years ago | og_kalu | venturebeat.com | newest

2

Causal Reasoning and Large Language Models: Opening a New Frontier for Causality (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

1

Demo/Weights out for Deepfloyd, SOTA(beats Imagen/Parti on FID) image generator (huggingface.co)

2 years ago | og_kalu | huggingface.co | newest

2

GPTs are Predictors, not Imitators or Simulators (alignmentforum.org)

2 years ago | og_kalu | alignmentforum.org | newest

1

Bing Has a Conversation with Cleverbot (reddit.com)

2 years ago | og_kalu | reddit.com | newest

1

GPTs are Predictors, not Imitators (lesswrong.com)

2 years ago | og_kalu | lesswrong.com | newest

4

GPT-4 has its own compression language (twitter.com/mckaywrigley)

2 years ago | og_kalu | twitter.com | newest

1

Large Language Models for Machine Translation (github.com/ogkalu2)

2 years ago | og_kalu | github.com | newest

13

ChemCrow: Augmenting large-language models with chemistry tools (arxiv.org)

2 years ago | og_kalu | arxiv.org | frontpage

5

Kandinsky 2.1: open-source txt2img generator, image blending, beats SD on FID (github.com/ai-forever)

2 years ago | og_kalu | github.com | newest

1

Large Language Models perform autonomous scientific research (paperswithcode.com)

2 years ago | og_kalu | paperswithcode.com | newest

2

Emergent autonomous scientific research capabilities of large language models (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest

1

Generative AI for Law, Harvey signs deals with some of the largest law firms (twitter.com/ai__pub)

2 years ago | og_kalu | twitter.com | newest

1

Generative AI is dreaming up new proteins (acs.org)

2 years ago | og_kalu | acs.org | newest

5

With LLMs, Researchers create Generative agents that interact with each other (twitter.com/nonmayorpete)

2 years ago | og_kalu | twitter.com | newest

37

Stanford benchmarks and compares numerous Large Language Models (stanford.edu)

2 years ago | og_kalu | stanford.edu | best

1

Humans in Humans Out: GPT Converging Toward Common Sense in Success and Failure (arxiv.org)

2 years ago | og_kalu | arxiv.org | newest