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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

letico

-0.72

ovo

-0.65

headers

-0.65

 interven

-0.64

partisan

-0.62

iot

-0.61

lez

-0.61

nikov

-0.60

Repe

-0.60

ividual

-0.59

POSITIVE LOGITS

 considered

0.99

 still

0.97

still

0.83

new

0.82

new

0.73

 rumored

0.71

New

0.69

 deemed

0.68

 unheard

0.63

isters

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.