INDEX

Explanations

dates formatted in year ranges

oai_token-act-pair · gpt-3.5-turbo

numerical data or values associated with events or statistics

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.74

 Barron

-0.69

 Morrison

-0.69

 Wright

-0.66

-0.65

resp

-0.63

asant

-0.62

rup

-0.62

awa

-0.62

emort

-0.61

POSITIVE LOGITS

0.93

0.90

0.84

0.82

bley

0.80

0.75

ension

0.74

endish

0.70

0.69

 sixteen

0.69

Activations Density 0.156%

dates formatted in year ranges

numerical data or values associated with events or statistics

No Comments

No Known Activations