INDEX

Explanations

phrases indicating support or opposition to various issues or individuals

oai_token-act-pair · gpt-3.5-turbo

phrases expressing support or endorsement

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Shant

-0.65

 earthqu

-0.65

 Suite

-0.65

lear

-0.63

pox

-0.63

worms

-0.62

glass

-0.62

chemy

-0.62

Recipe

-0.61

 burg

-0.61

POSITIVE LOGITS

votes

0.94

 endorse

0.93

 endorsing

0.92

Support

0.92

 Supports

0.89

itism

0.88

vote

0.86

enance

0.83

 endorsements

0.78

Supporters

0.76

Activations Density 0.073%

phrases indicating support or opposition to various issues or individuals

phrases expressing support or endorsement

No Comments

No Known Activations