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

bum

-0.65

 Holo

-0.61

 maiden

-0.58

riel

-0.57

ocial

-0.57

 outdoors

-0.57

)</

-0.56

[_

-0.56

iew

-0.54

 timid

-0.54

POSITIVE LOGITS

ioned

0.87

pees

0.75

isson

0.72

 Meaning

0.71

llah

0.69

zers

0.68

alties

0.66

andem

0.66

strom

0.66

rons

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.