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

 Carbuncle

-0.73

 apocalypse

-0.67

iesel

-0.66

geoning

-0.65

 Bolt

-0.64

 occult

-0.64

 suicidal

-0.63

 Pagan

-0.62

 wheelchair

-0.62

 coma

-0.62

POSITIVE LOGITS

fw

0.82

FW

0.78

dn

0.76

MEN

0.75

ESSION

0.74

Cmd

0.73

WER

0.73

window

0.73

tc

0.70

LE

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.