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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

eger

-0.77

ãĤ´ãĥ³

-0.75

ãĤ¦ãĤ¹

-0.75

monton

-0.70

AFB

-0.70

 coined

-0.70

 à¨

-0.63

borgh

-0.62

agus

-0.61

ãĥĥãĥī

-0.60

POSITIVE LOGITS

 sham

0.73

JO

0.68

tp

0.63

nox

0.61

wi

0.61

Iv

0.60

akov

0.59

 recess

0.59

tub

0.58

 Zoro

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.