INDEX

Explanations

phrases related to multiple storylines or plot elements within a narrative

oai_token-act-pair · gpt-3.5-turbo

references to narrative structures and storylines

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

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

cot

-0.89

ulia

-0.77

cham

-0.76

ury

-0.74

vil

-0.72

gn

-0.72

negie

-0.71

unte

-0.71

 Precision

-0.71

vard

-0.69

POSITIVE LOGITS

 storylines

1.11

 storyline

1.09

 backstory

1.02

acters

1.00

 flashbacks

0.99

 arcs

0.99

 protagonist

0.95

 heroine

0.95

 flashback

0.95

 narration

0.94

Activations Density 0.125%

phrases related to multiple storylines or plot elements within a narrative

references to narrative structures and storylines

No Comments

No Known Activations