INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ascus

-0.82

shock

-0.72

omething

-0.71

osterone

-0.71

dn

-0.65

Cyn

-0.65

 Lowell

-0.64

å£«

-0.61

emale

-0.61

HRC

-0.61

POSITIVE LOGITS

rez

0.81

 subdiv

0.66

--------------------------------------------------------

0.64

 Towns

0.64

 legisl

0.63

 vigilant

0.62

 cleaners

0.62

riend

0.62

////////

0.62

 Parties

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.