INDEX

Explanations

instances of the word "once" and its variations, indicating sequences or conditions in a narrative

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

after or once followed by a state

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

styleType

-0.50

cipais

-0.44

RTHOOK

-0.43

 particularly

-0.41

auer

-0.41

RefNanny

-0.40

NK

-0.40

leste

-0.40

nk

-0.40

tymologie

-0.39

POSITIVE LOGITS



0.49

posedge

0.45

abestanden

0.42

存于互联网档案馆

0.40

 jälkeen

0.40

 wikipagina

0.39

addContainerGap

0.39

).]

0.39

 Procedure

0.38

weile

0.38

Activations Density 0.029%

instances of the word "once" and its variations, indicating sequences or conditions in a narrative

after or once followed by a state

No Comments

No Known Activations