INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

George

-0.66

 Sting

-0.62

âĢ¢âĢ¢âĢ¢âĢ¢

-0.62

eday

-0.62

Â¯Â¯Â¯Â¯

-0.60

Rat

-0.60

 Orig

-0.59

 Drift

-0.59

 Presbyter

-0.59

worldly

-0.58

POSITIVE LOGITS

daq

0.91

cale

0.84

amia

0.78

neau

0.75

 monop

0.68

ription

0.68

aste

0.66

 eSports

0.66

uador

0.66

hua

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.