INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

llah

-0.80

HCR

-0.71

Disc

-0.68

 Defin

-0.66

Tweet

-0.65

Leaks

-0.64

 Vine

-0.62

 dissent

-0.62

 Share

-0.62

Ø§Ø

-0.61

POSITIVE LOGITS

enegger

0.85

0.81

agonist

0.80

ministic

0.77

abetic

0.76

ãĤ´ãĥ³

0.72

 Normandy

0.70

genic

0.68

ãĥ¼ãĥ³

0.68

idan

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.