INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Borough

-0.78

 Vessel

-0.71

 sided

-0.66

 Planes

-0.64

 stadiums

-0.62

 fined

-0.62

 Sold

-0.62

 Prin

-0.62

 Taxi

-0.61

 parach

-0.59

POSITIVE LOGITS

awar

0.73

SPONSORED

0.73

law

0.73

typ

0.72

papers

0.69

ACTION

0.69

mit

0.69

AppData

0.68

keleton

0.68

ãĤ£

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.