INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

âĸ¬

-0.84

emetery

-0.81

oway

-0.80

oice

-0.80

uria

-0.78

igue

-0.78

veyard

-0.77

hester

-0.77

ramid

-0.76

INGTON

-0.74

POSITIVE LOGITS

respond

0.70

 Emin

0.69

 Brus

0.68

ienne

0.68

Sof

0.66

 Melania

0.65

fac

0.65

naires

0.64

POR

0.63

zes

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.