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

urses

-0.75

itas

-0.70

urst

-0.69

 assemblies

-0.68

ies

-0.65

 mathematic

-0.61

OE

-0.61

oes

-0.60

Cmd

-0.60

socket

-0.59

POSITIVE LOGITS

egal

0.76

Quote

0.73

advertising

0.73

 Flavoring

0.72

idious

0.70

clave

0.69

>"

0.69

ktop

0.66

 ..."

0.65

rection

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.