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

gamer

-0.82

 à¨

-0.78

ãĤ¬

-0.78

externalActionCode

-0.77

ãĤ®

-0.75

ESA

-0.73

ãĥ¼ãĤ¯

-0.72

chell

-0.72

estamp

-0.71

AMES

-0.71

POSITIVE LOGITS

is

0.77

 divides

0.67

 Nina

0.64

 Watts

0.64

 controversy

0.63

 Goldberg

0.62

 agrees

0.62

ield

0.61

 organ

0.61

 floats

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.