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

tenance

-0.86

wagen

-0.61

 systems

-0.61

CAR

-0.60

Released

-0.59

Dod

-0.58

 stabilization

-0.57

 rebuild

-0.57

 fixtures

-0.57

 maintenance

-0.57

POSITIVE LOGITS

urtles

0.71

 Cosponsors

0.71

ixt

0.68

iq

0.68

esides

0.67

obar

0.66

ymes

0.64

owship

0.64

bidden

0.63

geries

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.