@_akhaliq - AI Dynamics - Page 129 of 138

Grounding Instructional Steps in Narrated How-To Videos

By

–

07 June 2023 4h14

Learning to Ground Instructional Articles in Videos through Narrations paper page: https://
huggingface.co/papers/2306.03
802
… present an approach for localizing steps of procedural activities in narrated how-to videos. To deal with the scarcity of labeled data at scale, we source the step

→ View original post on X — @_akhaliq

7 June 2023

Emergent Correspondence Discovery in Image Diffusion Models

By

@_akhaliq

–

07 June 2023 3h56

Emergent Correspondence from Image Diffusion paper page: https://
huggingface.co/papers/2306.03
881
… Finding correspondences between images is a fundamental problem in computer vision. In this paper, we show that correspondence emerges in image diffusion models without any explicit supervision. We

→ View original post on X — @_akhaliq

7 June 2023

Mega-TTS: Zero-Shot Text-to-Speech Scaling with Inductive Bias

By

@_akhaliq

–

07 June 2023 3h28

Mega-TTS: Zero-Shot Text-to-Speech at Scale with Intrinsic Inductive Bias paper page: https://
huggingface.co/papers/2306.03
509
… Scaling text-to-speech to a large and wild dataset has been proven to be highly effective in achieving timbre and speech style generalization, particularly in zero-shot

→ View original post on X — @_akhaliq

7 June 2023

LEACE: Perfect Linear Concept Erasure in Closed Form

By

@_akhaliq

–

07 June 2023 3h27

LEACE: Perfect linear concept erasure in closed form paper page: https://
huggingface.co/papers/2306.03
819
… introduce LEAst-squares Concept Erasure (LEACE), a closed-form method which provably prevents all linear classifiers from detecting a concept while inflicting the least possible damage to

→ View original post on X — @_akhaliq

7 June 2023

Natural Language Commanding via Program Synthesis for Productivity Software

By

@_akhaliq

–

07 June 2023 3h25

Natural Language Commanding via Program Synthesis paper page: https://
huggingface.co/papers/2306.03
460
… present Semantic Interpreter, a natural language-friendly AI system for productivity software such as Microsoft Office that leverages large language models (LLMs) to execute user intent across

→ View original post on X — @_akhaliq

7 June 2023

Large Language Models Struggle With Code Containing Potential Bugs

By

@_akhaliq

–

07 June 2023 3h23

Large Language Models of Code Fail at Completing Code with Potential Bugs paper page: https://
huggingface.co/papers/2306.03
438
… Large language models of code (Code-LLMs) have recently brought tremendous advances to code completion, a fundamental feature of programming assistance and code

→ View original post on X — @_akhaliq

7 June 2023

Static Evaluation of Code Completion by Large Language Models

By

@_akhaliq

–

07 June 2023 3h18

A Static Evaluation of Code Completion by Large Language Models paper page: https://
huggingface.co/papers/2306.03
203
… Large language models trained on code have shown great potential to increase productivity of software developers. Several execution-based benchmarks have been proposed to

→ View original post on X — @_akhaliq

7 June 2023

Deductive Verification of Chain-of-Thought Reasoning in Large Language Models

By

@_akhaliq

–

07 June 2023 3h16

Deductive Verification of Chain-of-Thought Reasoning paper page: https://
huggingface.co/papers/2306.03
872
… Large Language Models (LLMs) significantly benefit from Chain-of-Thought (CoT) prompting in performing various reasoning tasks. While CoT allows models to produce more comprehensive

→ View original post on X — @_akhaliq

7 June 2023