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

 GOODMAN

-0.64

ustomed

-0.63

 Aval

-0.63

 relayed

-0.62

pez

-0.61

iants

-0.60

 Takeru

-0.60

LINE

-0.59

uality

-0.58

ouf

-0.57

POSITIVE LOGITS

Ü

0.76

 depreciation

0.75

âķĲâķĲ

0.70

okÃ©

0.68

-+

0.67

hay

0.66

ste

0.65

 scrut

0.65

ueller

0.64

 probation

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.