Stories by og_kalu

GPT-4 frustrated with failing browsing tool, searches how to switch the browser

Evidence of Meaning in Large Language Models Trained on Programs

MusicGen: Simple and controllable music generation

Experimental results from applying GPT-4 to an unpublished formal language

StyleDrop: Text-to-Image Generation in Any Style

LLM Itself Can Read and Generate CXR Images

Fine-Tuning Language Models with Just Forward Passes

A Mechanistic Interpretability Analysis of Grokking

Tiny Transformer trained for addition and examined with bizarre results

Improving Factuality and Reasoning in Language Models Through Multiagent Debate

RecurrentGPT: Interactive Generation of (Arbitrarily) Long Text

Bing vs Cleverbot

Evidence of Meaning in Language Models Trained on Programs

Tree of Thoughts: Deliberate Problem Solving with Large Language Models

SoundStorm: Efficient Parallel Audio Generation. 30s dialogue generated in 2s

Palm 2 – a 340B model trained on 3.6 Trillion tokens of text

Symbol tuning improves in-context learning in language models

TinyStories: How Small Can Language Models Be and Still Speak Coherent English?

Code trained LLMs reason better, on benchmarks that have nothing to do with code

TidyBot: Personalized Robot Assistance with LLMs

4
5
6
7
8
9