INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

shift

-0.73

FML

-0.72

externalActionCode

-0.72

UMP

-0.72

control

-0.72

ctrl

-0.72

tein

-0.71

Gaza

-0.69

-0.68

-0.65

POSITIVE LOGITS

asions

0.73

æ³

0.71

 Rebell

0.68

 Budd

0.65

orative

0.64

Vaj

0.64

haps

0.63

 invitations

0.62

 comparisons

0.62

 attm

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.