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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

jog

-0.75

 descended

-0.74

leon

-0.70

Demon

-0.65

hunt

-0.65

demon

-0.63

gdala

-0.62

 greeted

-0.62

 conver

-0.62

 ALWAYS

-0.61

POSITIVE LOGITS

Sov

0.84

 ceilings

0.75

skin

0.69

 Consent

0.68

SpaceEngineers

0.67

 Salary

0.66

arez

0.65

achev

0.64

elin

0.64

Pri

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.