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

¥µ

-0.77

igent

-0.72

 Gathering

-0.71

 Aware

-0.66

Ĥ¬

-0.63

 catalog

-0.59

 Tire

-0.59

Î³

-0.59

idable

-0.58

 Catalog

-0.58

POSITIVE LOGITS

[_

0.72

selection

0.71

acters

0.69

 Peninsula

0.68

olean

0.66

oat

0.64

riott

0.64

rooms

0.63

 peninsula

0.63

======

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.