INDEX

Explanations

references to things being given or provided

oai_token-act-pair · gpt-3.5-turbo

phrases indicating direction or recipients of actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

natureconservancy

-0.86

tions

-0.64

steen

-0.64

hov

-0.64

Gap

-0.59

ierre

-0.59

 endorsements

-0.56

gone

-0.56

 beginnings

-0.55

NOT

-0.55

POSITIVE LOGITS

ãĥīãĥ©

0.81

å§«

0.80

by

0.79

ãĤ¨ãĥ«

0.78

ãĥĨãĤ£

0.74

Ń·

0.73

 commemorate

0.73

que

0.72

BY

0.72

estone

0.72

Activations Density 0.249%

references to things being given or provided

phrases indicating direction or recipients of actions

No Comments

No Known Activations