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

Negative Logits

¿½

-0.90

ĸļ

-0.80

 Meadows

-0.77

 weave

-0.69

 quar

-0.68

 bends

-0.67

oma

-0.67

ĨĴ

-0.66

ijk

-0.66

©¶æ

-0.65

POSITIVE LOGITS

rolet

0.87

sys

0.71

hip

0.67

visors

0.66

Delivery

0.65

livious

0.65

Berry

0.64

bodied

0.64

ILY

0.63

payers

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.