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

éĸ

-0.85

å£«

-0.73

Recomm

-0.72

 confir

-0.72

ä½¿

-0.71

 Rohing

-0.71

DIT

-0.67

å·

-0.66

 Intermediate

-0.65

PLIED

-0.64

POSITIVE LOGITS

ipl

0.73

kefeller

0.70

 cloves

0.63

selves

0.63

idine

0.62

 knots

0.62

seq

0.62

ulence

0.61

ificate

0.61

wine

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.