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

APH

-0.73

SI

-0.71

Interstitial

-0.70

PE

-0.70

MAS

-0.69

INA

-0.68

apore

-0.67

chen

-0.65

 Palest

-0.65

anian

-0.64

POSITIVE LOGITS

0.89

wt

0.74

vernment

0.72

wa

0.64

 reliance

0.64

 Witness

0.63

ieth

0.62

wat

0.60

irmation

0.60

 biod

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.