INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

awa

-0.72

 literacy

-0.62

])

-0.62

chin

-0.61

>>>>>>>>

-0.61

 rapport

-0.61

eston

-0.60

 ardent

-0.60

ief

-0.60

 Forth

-0.59

POSITIVE LOGITS

abouts

0.74

ruciating

0.72

tions

0.72

 gadgets

0.69

aptic

0.68

istries

0.67

 Comet

0.67

utions

0.66

ionic

0.66

tz

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.