INDEX

Explanations

words related to physical actions or tasks

oai_token-act-pair · gpt-3.5-turbo

specific single-letter or two-letter words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Ô

-1.00

FUL

-0.95

stellar

-0.80

cipled

-0.74

imentary

-0.72

 Presbyterian

-0.71

 Continued

-0.71

landish

-0.70

)=(

-0.68

theless

-0.66

POSITIVE LOGITS

asers

0.94

icion

0.92

glers

0.91

ancies

0.90

akers

0.89

arers

0.89

aunts

0.88

ags

0.87

ippers

0.86

amps

0.86

Activations Density 0.232%

words related to physical actions or tasks

specific single-letter or two-letter words

No Comments

No Known Activations