INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

aunts

-0.65

seed

-0.63

emer

-0.62

icted

-0.62

hoe

-0.61

 Peninsula

-0.60

untled

-0.60

SpaceEngineers

-0.59

ocument

-0.59

Enlarge

-0.59

POSITIVE LOGITS

ĸļ

0.99

roth

0.71

uckland

0.68

ctic

0.68

ometry

0.66

kus

0.66

acci

0.66

olate

0.66

onement

0.66

 nutshell

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.