INDEX

Explanations

details related to actions or events that happen before a specific event

oai_token-act-pair · gpt-3.5-turbo

instances of the word "before."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

geist

-0.82

aez

-0.80

OTOS

-0.73

hack

-0.71

ãĥŃ

-0.70

orage

-0.70

uilt

-0.69

Fight

-0.69

arta

-0.68

illed

-0.68

POSITIVE LOGITS

 settling

1.34

 disappearing

1.33

 returning

1.27

 concluding

1.21

 deciding

1.20

 retreating

1.20

 vanishing

1.19

 turning

1.18

 reaching

1.17

 retiring

1.17

Activations Density 0.055%

details related to actions or events that happen before a specific event

instances of the word "before."

No Comments

No Known Activations