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

ãĥ¼ãĥĨãĤ£

-0.75

icating

-0.71

 alley

-0.63

undy

-0.62

ication

-0.62

 Mare

-0.62

icators

-0.62

 cooler

-0.61

agra

-0.60

icator

-0.60

POSITIVE LOGITS

DEC

0.84

NOT

0.82

EV

0.79

EMP

0.75

payer

0.75

placed

0.74

EG

0.73

Ber

0.73

cards

0.70

LIST

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.