INDEX

Explanations

words related to controversial topics or statements

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter 'b'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Wer

-0.60

 BART

-0.59

bilt

-0.58

 Stall

-0.58

 Audi

-0.57

hyd

-0.57

 phon

-0.56

 justice

-0.56

 Bayern

-0.56

 tunes

-0.55

POSITIVE LOGITS

rought

1.43

rief

1.31

odies

1.30

esides

1.28

izarre

1.26

ombs

1.26

urden

1.25

ureau

1.21

eware

1.19

usting

1.18

Activations Density 0.037%

words related to controversial topics or statements

occurrences of the letter 'b'

No Comments

No Known Activations