TiDAR: Think in Diffusion, Talk in Autoregression (arxiv.org)

Screenshot of arxiv.org: TiDAR: Think in Diffusion, Talk in Autoregression

First screen, captured 11 months ago.

Looks similar

Screenshot of arxiv.org: Neural Network Diffusion

Neural Network Diffusion (arxiv.org)

199 points by vagabund 2 years ago | 74 comments

Screenshot of arxiv.org: DiffusionGemma Technical Report

DiffusionGemma Technical Report (arxiv.org)

146 points by gmays 2 months ago | 35 comments

Screenshot of arxiv.org: Scalable MatMul-Free Language Modeling

Scalable MatMul-Free Language Modeling (arxiv.org)

177 points by lykahb 2 years ago | 27 comments

Screenshot of arxiv.org: Teuken-7B-Base and Teuken-7B-Instruct: Towards European LLMs

Teuken-7B-Base and Teuken-7B-Instruct: Towards European LLMs (arxiv.org)

235 points by doener 1 year ago | 88 comments

Screenshot of arxiv.org: Microsoft says GPT 3.5 has 20B parameters?

Microsoft says GPT 3.5 has 20B parameters?(arxiv.org)

44 points by Heidaradar 3 years ago | 11 comments

Screenshot of arxiv.org: Differential Transformer

Differential Transformer (arxiv.org)

525 points by weirdcat 2 years ago | 166 comments

arXiv paper introducing Diff Transformer, which computes attention as the difference of two softmax maps to cancel noise and improve long-context modeling and reduce hallucination.

differential-transformer attention-mechanism transformers large-language-models

Looks similar

Neural Network Diffusion(arxiv.org)

DiffusionGemma Technical Report(arxiv.org)

Scalable MatMul-Free Language Modeling(arxiv.org)

Teuken-7B-Base and Teuken-7B-Instruct: Towards European LLMs(arxiv.org)