search
Classify:AIIt Industry

From RLHF to DPO to TDPO, large model alignment algorithms are already 'token-level'

Release:2024-06-24 15:04:43
From RLHF to DPO to TDPO, large model alignment algorithms are already 'token-level'

Why have you given up on LangChain?

Release:2024-06-24 14:20:32
Why have you given up on LangChain?

ICML 2024 | Feature pollution: Neural networks learn irrelevant features and fail to generalize

Release:2024-06-24 14:17:11
ICML 2024 | Feature pollution: Neural networks learn irrelevant features and fail to generalize

CVPR 24|ETH Zurich and other teams: Redefining the small sample 3D segmentation task, the new benchmark opens up broad potential for improvement!

Release:2024-06-24 14:05:51
CVPR 24|ETH Zurich and other teams: Redefining the small sample 3D segmentation task, the new benchmark opens up broad potential for improvement!

ByteDouBao's new image Tokenizer: only 32 tokens are needed to generate an image, and the speed is increased by up to 410 times.

Release:2024-06-24 14:03:31
ByteDouBao's new image Tokenizer: only 32 tokens are needed to generate an image, and the speed is increased by up to 410 times.

Huawei Cloud uses Pangu large model to reshape services, and ModelArts Studio is fully upgraded

Release:2024-06-22 11:55:38
Huawei Cloud uses Pangu large model to reshape services, and ModelArts Studio is fully upgraded

First-hand review of Claude 3.5: Is it really better than GPT-4o for playing tricks, seeing a doctor, playing tricks, and doing math problems?

Release:2024-06-22 07:46:19
First-hand review of Claude 3.5: Is it really better than GPT-4o for playing tricks, seeing a doctor, playing tricks, and doing math problems?

Covering text, positioning and segmentation tasks, Zhiyuan and Hong Kong Chinese jointly proposed the first multi-functional 3D medical multi-modal large model

Release:2024-06-22 07:16:49
Covering text, positioning and segmentation tasks, Zhiyuan and Hong Kong Chinese jointly proposed the first multi-functional 3D medical multi-modal large model

Whether it's real or AI video, 'Mosca” can reconstruct and restore 4D dynamic renderable scenes.

Release:2024-06-22 07:09:52
Whether it's real or AI video, 'Mosca” can reconstruct and restore 4D dynamic renderable scenes.

Automatically identify the best molecules and reduce synthesis costs. MIT develops a molecular design decision-making algorithm framework

Release:2024-06-22 06:43:34
Automatically identify the best molecules and reduce synthesis costs. MIT develops a molecular design decision-making algorithm framework

WAIC Forum Registration|Gathering together many central state-owned enterprises and AI companies to discuss the integrated development of industries

Release:2024-06-22 04:59:29
WAIC Forum Registration|Gathering together many central state-owned enterprises and AI companies to discuss the integrated development of industries

Xiaoice and ShiMeta co-create ShiMeta digital human solution: with the blessing of large models, every large-screen terminal becomes a new interactive carrier

Release:2024-06-22 02:22:30
Xiaoice and ShiMeta co-create ShiMeta digital human solution: with the blessing of large models, every large-screen terminal becomes a new interactive carrier

Can machine learning be done purely by talking without doing numerical calculations? A new ML paradigm based on natural language is coming

Release:2024-06-22 02:06:40
Can machine learning be done purely by talking without doing numerical calculations? A new ML paradigm based on natural language is coming

Beat GPT-4o! The new king Claude 3.5 Sonnet is here, available directly for free

Release:2024-06-22 01:08:30
Beat GPT-4o! The new king Claude 3.5 Sonnet is here, available directly for free

Luma and Runway take turns to explode, and video generation reaches new heights. Can Sora still dominate? Come to this WAIC video generation forum to find answers

Release:2024-06-21 22:00:23
Luma and Runway take turns to explode, and video generation reaches new heights. Can Sora still dominate? Come to this WAIC video generation forum to find answers