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

Ħ¢

-0.83

Interstitial

-0.83

ahon

-0.74

rama

-0.73

allery

-0.72

osta

-0.72

OTT

-0.71

raz

-0.69

 corrid

-0.69

querque

-0.69

POSITIVE LOGITS

Ps

0.66

 Guant

0.66

Geh

0.65

 Gates

0.63

NSA

0.62

Sno

0.61

 Founding

0.60

ĸļ

0.59

 Dusk

0.58

pal

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.