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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 rode

-0.78

 livest

-0.76

lived

-0.69

vertisement

-0.68

sat

-0.64

arious

-0.63

acebook

-0.63

nz

-0.62

esses

-0.62

rl

-0.61

POSITIVE LOGITS

 Jungle

0.77

 Morocco

0.70

Reloaded

0.68

 Archdemon

0.66

AAF

0.65

tery

0.65

 Ankara

0.65

 Afric

0.63

 Mane

0.63

 Optional

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.