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

uality

-0.87

manship

-0.75

 Plex

-0.67

 duplication

-0.67

jew

-0.63

 Hardcore

-0.63

ijn

-0.63

 altogether

-0.61

ivism

-0.60

ĺħ

-0.59

POSITIVE LOGITS

anta

0.92

URI

0.70

estate

0.66

ocating

0.65

iott

0.62

avorite

0.62

 reluctant

0.60

 trending

0.60

 notified

0.59

eatures

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.