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

earchers

-1.00

reditary

-0.97

imore

-0.92

htaking

-0.91

mares

-0.88

naire

-0.86

ritic

-0.86

iseum

-0.85

itary

-0.83

nesota

-0.82

POSITIVE LOGITS

 limit

0.72

take

0.67

Shape

0.66

 Consent

0.64

 Archangel

0.63

ÂłÂł

0.63

 lett

0.60

Initialized

0.59

 Intercept

0.59

 Lois

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.