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

mins

-0.68

 Scrib

-0.66

 Rollins

-0.65

ogical

-0.65

anse

-0.63

 Cheong

-0.61

 Prec

-0.61

 Wrest

-0.60

 Thou

-0.60

 Seah

-0.60

POSITIVE LOGITS

indal

0.76

NV

0.71

maxwell

0.64

enf

0.63

hra

0.63

wcsstore

0.63

Motor

0.63

Border

0.63

mares

0.60

GV

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.