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

sticks

-0.64

fig

-0.63

 âĵĺ

-0.61

 Awoken

-0.60

vis

-0.59

ado

-0.59

 sticks

-0.58

-0.57

atana

-0.57

AV

-0.57

POSITIVE LOGITS

glomer

0.73

zl

0.73

aples

0.73

infeld

0.70

etimes

0.69

urus

0.69

0.68

translation

0.66

everal

0.66

ymes

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.