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

 embr

-0.65

icators

-0.64

icator

-0.64

eller

-0.63

 Orient

-0.63

ection

-0.62

pleting

-0.61

azines

-0.61

save

-0.61

agna

-0.60

POSITIVE LOGITS

twitter

0.84

 lately

0.69

FACE

0.69

UST

0.68

stro

0.67

BRE

0.67

ocy

0.67

staff

0.67

æĦ

0.66

NET

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.