INDEX

Explanations

proper nouns with historical and political significance

oai_token-act-pair · gpt-3.5-turbo

references to historical political entities and movements

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

maxwell

-0.80

ebin

-0.78

 Saying

-0.74

 saying

-0.71

 tweeting

-0.66

Write

-0.65

 Kardash

-0.64

\":

-0.64

 GoPro

-0.63

 WATCH

-0.63

POSITIVE LOGITS

 flourished

1.52

 consisted

1.44

 lasted

1.41

 became

1.35

 existed

1.34

 differed

1.30

 ceased

1.30

 remained

1.29

 persisted

1.28

 tended

1.27

Activations Density 0.495%

proper nouns with historical and political significance

references to historical political entities and movements

No Comments

No Known Activations