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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ASC

-0.75

âĢ¢âĢ¢

-0.70

 Syndicate

-0.66

 Academy

-0.64

ISI

-0.64

 nuisance

-0.64

 plurality

-0.63

 Integrity

-0.62

 Mustang

-0.62

Moh

-0.62

POSITIVE LOGITS

acon

0.77

rower

0.75

uel

0.72

yrus

0.72

artifacts

0.70

riott

0.70

leeve

0.69

mort

0.67

herty

0.67

ommel

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.