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

 concess

-0.76

andem

-0.73

endi

-0.68

gem

-0.67

oru

-0.66

irs

-0.65

bats

-0.65

ollen

-0.64

ouf

-0.64

leck

-0.64

POSITIVE LOGITS

çĶ°

0.93

å¥³

0.83

ossibility

0.77

votes

0.72

actionDate

0.71

 Rite

0.69

¶ħ

0.69

cir

0.66

Ñĭ

0.64

essee

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.