INDEX

Explanations

specific named entities or proper nouns in the text

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

Don Gibson

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

tanleria

-0.44

 AppComponent

-0.43

providedIn

-0.43

PerformLayout

-0.41

kmäler

-0.41

 OMITTED

-0.40

 المعيارى

-0.40

Gemeinden

-0.39

UrlResolution

-0.38

 насељу

-0.38

POSITIVE LOGITS

yntaxException

0.53

 betweenstory

0.48

zyst

0.43

ValueStyle

0.43

tako

0.42

InjectAttribute

0.40

 Salam

0.39

ammans

0.38

Derbyniad

0.38

addContainerGap

0.37

Activations Density 0.095%

specific named entities or proper nouns in the text

Don Gibson

No Comments

No Known Activations