INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ecast

-0.81

alg

-0.79

ouls

-0.73

sonian

-0.71

pring

-0.70

ibrary

-0.69

berries

-0.66

fall

-0.65

url

-0.63

Ur

-0.63

POSITIVE LOGITS

iform

0.66

 Solutions

0.61

 Sector

0.60

 Petty

0.60

ATING

0.60

senal

0.59

ISION

0.58

 Difference

0.57

GROUP

0.57

GBT

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.