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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 juices

-0.79

cules

-0.72

 lucid

-0.70

atsu

-0.69

inous

-0.69

chant

-0.66

emon

-0.66

umbers

-0.66

esome

-0.65

tap

-0.65

POSITIVE LOGITS

cru

0.76

 Meaning

0.73

NYSE

0.66

 clipboard

0.65

 Were

0.64

peria

0.64

 Cipher

0.63

 Photography

0.62

Cru

0.61

Berry

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.