INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

urers

-0.86

culosis

-0.79

REF

-0.78

perty

-0.78

ishly

-0.73

ortium

-0.72

tenance

-0.71

prus

-0.70

notations

-0.68

tyard

-0.67

POSITIVE LOGITS

chan

0.68

 Holo

0.66

lip

0.65

 Appears

0.61

olla

0.61

McA

0.61

 Fork

0.61

central

0.60

Aut

0.60

 =================================================================

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.