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

wash

-0.69

enegger

-0.66

ruby

-0.65

 parade

-0.63

 bearer

-0.62

given

-0.62

Ô

-0.60

adesh

-0.59

pee

-0.59

bred

-0.58

POSITIVE LOGITS

 sarc

0.66

nep

0.65

Ear

0.59

NRS

0.58

 Hots

0.57

unct

0.57

imar

0.56

 Rent

0.56

ortium

0.56

 turb

0.55

Activations Density 0.000%

No Known Activations

This feature has no known activations.