INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

leck

-0.83

Neal

-0.82

 ç¥ŀ

-0.74

 Weasley

-0.64

 punch

-0.64

ueller

-0.63

 flaw

-0.61

 Exploration

-0.61

arnaev

-0.61

weights

-0.60

POSITIVE LOGITS

conn

0.76

Ļ½

0.68

atri

0.66

entry

0.66

roc

0.65

rav

0.64

ģĸ

0.63

tro

0.63

pipe

0.62

 neglected

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.