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

iat

-0.79

pos

-0.77

 WATCHED

-0.73

alyses

-0.69

nis

-0.69

nav

-0.69

inez

-0.68

meter

-0.67

wait

-0.66

scan

-0.65

POSITIVE LOGITS

 Clyde

0.77

OULD

0.72

 ensu

0.70

 Isles

0.64

clud

0.64

 defe

0.63

 Scotland

0.61

wee

0.61

 Robbie

0.60

eld

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.