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

Cec

-0.71

sic

-0.69

Ore

-0.68

Cooldown

-0.66

DAQ

-0.65

CG

-0.64

 cosmetics

-0.64

 toile

-0.62

isable

-0.62

ties

-0.61

POSITIVE LOGITS

oping

0.75

 annex

0.74

dinand

0.74

ewitness

0.70

renheit

0.69

ushima

0.66

ueller

0.63

unks

0.63

ocalyptic

0.63

 tread

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.