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

 Ames

-0.72

 Gemini

-0.70

 Category

-0.68

roleum

-0.62

ium

-0.62

Jaw

-0.61

 Nasa

-0.61

 astroph

-0.60

 Hubble

-0.60

 Volcano

-0.60

POSITIVE LOGITS

nces

0.81

 captcha

0.78

cia

0.77

ooks

0.76

ource

0.76

Tes

0.72

UX

0.68

ires

0.68

orthy

0.66

endar

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.