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

-0.79

security

-0.64

 Prot

-0.60

 Trend

-0.59

ï¿½ï¿½ï¿½ï¿½

-0.59

 maintaining

-0.59

Cour

-0.58

sec

-0.58

 maintained

-0.58

cas

-0.57

POSITIVE LOGITS

gered

0.78

acter

0.75

wra

0.72

minist

0.70

sie

0.70

iott

0.69

ratulations

0.69

umbnail

0.66

ionage

0.66

ursion

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.