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

liness

-0.80

 forgiven

-0.76

 Gord

-0.71

 plun

-0.71

SPONSORED

-0.65

 Bauer

-0.63

itous

-0.63

mental

-0.61

mentioned

-0.61

:]

-0.61

POSITIVE LOGITS

ems

0.78

outheast

0.78

uri

0.75

AMES

0.73

imens

0.71

IGHTS

0.69

earch

0.66

unic

0.66

oggles

0.65

eteria

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.