INDEX

Explanations

phrases where something is mentioned as being done or happening in the first place

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

pherd

-0.74

Tags

-0.64

md

-0.63

 Izan

-0.62

âĶģ

-0.62

âĢİ

-0.62

quished

-0.60

ourge

-0.58

rx

-0.58

issance

-0.57

POSITIVE LOGITS

 place

1.26

 instance

1.04

place

0.89

 millenn

0.85

 Place

0.83

 places

0.80

 incarnation

0.79

 iteration

0.79

 edition

0.78

 guise

0.78

Activations Density 0.048%

phrases where something is mentioned as being done or happening in the first place

No Comments

No Known Activations