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

ren

-0.69

rises

-0.65

rax

-0.63

rama

-0.62

 occurs

-0.62

estones

-0.61

rie

-0.60

red

-0.60

rophe

-0.58

 derby

-0.58

POSITIVE LOGITS

Org

0.88

olicy

0.86

GBT

0.85

ettings

0.79

CHQ

0.74

er

0.73

 Erit

0.70

 lawy

0.70

ktop

0.70

iewicz

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.