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

wine

-0.91

Discussion

-0.80

limits

-0.77

DragonMagazine

-0.75

wolves

-0.75

existence

-0.72

currency

-0.69

-0.68

improve

-0.66

 Smile

-0.66

POSITIVE LOGITS

Coy

0.79

kie

0.70

 Parsons

0.62

 Flores

0.61

 COMPLE

0.60

Khe

0.59

 possibly

0.59

ulla

0.58

 furthe

0.58

ab

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.