INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

xual

-0.87

emale

-0.85

endium

-0.84

ternity

-0.84

ĸļ

-0.84

adobe

-0.83

bleacher

-0.80

akening

-0.79

erity

-0.78

notations

-0.77

POSITIVE LOGITS

 Torres

0.69

Jiu

0.65

Cub

0.61

RO

0.60

itional

0.60

 remotely

0.60

Oper

0.59

lan

0.58

 deposit

0.58

 transit

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.