Jump to main content Jump to sidebar

Forums
Wiki

Log in
Sign up

Overview
Submissions
Comments

MysteryInc152

Microsoft 365 Copilot – your copilot for work

blogs.microsoft.com

Submitted by MysteryInc152 t3_11sxofy on March 16, 2023 at 3:43 PM in technology

6 comments

8

Language models can now teach themselves HOW to use tools (ie. any API) in real time and completely automated. When given a task, SLAPA knows to search for the API documentation and learn all the information. Then he create API calls. If they don't work, he learns from his mistake and tries again.

twitter.com

Submitted by MysteryInc152 t3_11hof5q on March 4, 2023 at 3:40 AM in singularity

No comments

0

[R] ChatGLM-6B - an open source 6.2 billion parameter Eng/Chinese bilingual LLM trained on 1T tokens, supplemented by supervised fine-tuning, feedback bootstrap, and RLHF. Runs on consumer grade GPUs

github.com

Submitted by MysteryInc152 t3_11utpud on March 18, 2023 at 5:01 PM in MachineLearning

49 comments

201

[R] EvoPrompting: Language models can create novel and effective deep neural architectures. These architectures are also able to outperform those designed by human experts (with few-shot prompting)

Submitted by MysteryInc152 t3_11f26ig on March 1, 2023 at 12:10 PM in MachineLearning

6 comments

10

Large language models generate functional protein sequences across diverse families

nature.com

Submitted by MysteryInc152 t3_11ckc8a on February 26, 2023 at 3:46 PM in singularity

17 comments

60

[R] Meta AI open sources new SOTA LLM called LLaMA. 65B version (trained on 1.4T tokens) is competitive with Chinchilla and Palm-540B. 13B version outperforms OPT and GPT-3 175B on most benchmarks.

Submitted by MysteryInc152 t3_11awp4n on February 24, 2023 at 5:21 PM in MachineLearning

No comments

2

[D] Toolformer implementation using only few-shot prompting

twitter.com

Submitted by MysteryInc152 t3_115x1it on February 19, 2023 at 12:47 AM in MachineLearning

23 comments

88

[R] Large language models generate functional protein sequences across diverse families

nature.com

Submitted by MysteryInc152 t3_11ckcxv on February 26, 2023 at 3:46 PM in MachineLearning

13 comments

137

[R] Microsoft introduce Kosmos-1, a Multimodal Large Language Model (MLLM) that can perceive general modalities, learn in context (i.e., few-shot), and follow instructions (i.e., zero-shot)

Submitted by MysteryInc152 t3_11e4w40 on February 28, 2023 at 12:30 PM in MachineLearning

40 comments

330

Decoding fMRI based brain activities and reconstructing images with accurate semantics and image features using diffusion model

twitter.com

Submitted by MysteryInc152 t3_yx71zm on November 16, 2022 at 9:38 PM in singularity

25 comments

147

MysteryInc152

Registered on May 18, 2020

t2_6ifa7497

Running Postmill