INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 fingert

-0.69

aeper

-0.69

Untitled

-0.63

 cumbers

-0.61

 voic

-0.61

vati

-0.61

uliffe

-0.60

 todd

-0.60

 princ

-0.58

 squeeze

-0.58

POSITIVE LOGITS

displayText

0.71

mobi

0.71

arez

0.69

 WARN

0.68

Correct

0.66

jury

0.65

ament

0.62

license

0.62

acies

0.61

ãĤ¹ãĥĪ

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.