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

 prove

-0.67

rils

-0.66

chin

-0.63

¤

-0.62

Âł

-0.59

 lifts

-0.59

 conceal

-0.59

 borrow

-0.58

ĸļ

-0.58

 overlook

-0.58

POSITIVE LOGITS

wards

0.80

actionDate

0.75

aster

0.72

pires

0.72

ivery

0.68

 mosqu

0.67

roads

0.67

urate

0.66

 adolesc

0.65

effic

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.