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

Footnote

-0.75

Dialogue

-0.75

 veter

-0.74

Plot

-0.68

 Coulter

-0.64

 Shepard

-0.64

TextColor

-0.64

åħī

-0.63

vor

-0.63

Ĭ±

-0.62

POSITIVE LOGITS

ependence

0.78

 Feder

0.73

 Bane

0.71

istries

0.66

epend

0.66

iculty

0.66

gas

0.60

 slack

0.58

 âĶľâĶĢâĶĢ

0.58

ema

0.56

Activations Density 0.000%

No Known Activations

This feature has no known activations.