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

ucky

-0.72

phia

-0.64

 Budget

-0.63

 brig

-0.61

osa

-0.61

etime

-0.60

 Birds

-0.60

 polic

-0.59

uity

-0.59

ament

-0.59

POSITIVE LOGITS

é¾įåĸļå£«

0.79

arnaev

0.76

hran

0.76

urses

0.75

Accessory

0.75

FLAG

0.73

ults

0.72

 Haku

0.70

ADRA

0.70

 unres

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.