INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 poaching

-0.68

 Mold

-0.66

 Preservation

-0.64

 antit

-0.64

 fung

-0.62

 Underworld

-0.62

 Daylight

-0.62

 Vision

-0.60

 Protection

-0.60

BBQ

-0.60

POSITIVE LOGITS

eus

0.85

edIn

0.83

eger

0.80

phe

0.79

0.77

inoa

0.75

pmwiki

0.72

isition

0.72

sch

0.71

orney

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.