INDEX

Explanations

sequences or versions of the word "sequence"

oai_token-act-pair · gpt-3.5-turbo

terms related to sequencing and sequential processes

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 intens

-0.65

 traveller

-0.65

 bearer

-0.63

 Bunny

-0.60

 flared

-0.59

Kaf

-0.57

 discouraging

-0.56

 uncond

-0.56

 Commodore

-0.56

 owners

-0.55

POSITIVE LOGITS

enced

1.51

ential

1.38

encer

1.38

encing

1.37

entially

1.37

ences

1.36

ester

1.29

encers

1.20

entials

1.12

etry

1.09

Activations Density 0.050%

sequences or versions of the word "sequence"

terms related to sequencing and sequential processes

No Comments

No Known Activations