INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

irit

-0.91

rient

-0.72

ften

-0.72

azeera

-0.72

encer

-0.71

ync

-0.69

uyomi

-0.66

akening

-0.66

issance

-0.66

ivil

-0.64

POSITIVE LOGITS

mate

0.69

mates

0.67

 prostitute

0.65

CLA

0.65

wcsstore

0.64

 Polk

0.63

 lengths

0.63

 cleaners

0.62

rants

0.61

 McCorm

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.