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

istle

-0.68

anta

-0.67

phrine

-0.64

ray

-0.64

nesday

-0.64

zie

-0.62

isin

-0.62

VIDIA

-0.61

rehensive

-0.60

 withd

-0.59

POSITIVE LOGITS

oter

0.75

eries

0.69

emic

0.68

ãĥ¼ãĤ¯

0.68

amples

0.65

selves

0.63

 tuber

0.62

ATIONS

0.60

 tracts

0.60

otaur

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.