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

 derog

-0.73

downs

-0.67

hler

-0.67

audio

-0.65

gae

-0.64

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

-0.64

 quotas

-0.63

 retaliation

-0.62

friends

-0.62

itism

-0.61

POSITIVE LOGITS

ancock

0.81

 Gutenberg

0.73

 Surviv

0.72

 Pradesh

0.71

ogly

0.70

ospace

0.70

 Awareness

0.69

phis

0.67

 profession

0.65

 Technician

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.