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

erry

-0.73

lore

-0.71

oug

-0.71

BACK

-0.70

UID

-0.68

\\\\\\\\

-0.67

jar

-0.66

raising

-0.66

ural

-0.66

Ãį

-0.63

POSITIVE LOGITS

ĻĤ

0.64

upt

0.63

=#

0.62

 Depression

0.60

oxide

0.58

 depression

0.58

 Levine

0.57

engeance

0.57

=""

0.57

pless

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.