INDEX

Explanations

braces or brackets used in programming syntax

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

return value or property

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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



-0.72

 queſta

-0.72

 صوتيه

-0.63

__(/*!

-0.62

Personendaten

-0.60

 endblock

-0.57

⦑

-0.57

WaitGroup

-0.56

 wireType

-0.56

withIdentifier

-0.56

POSITIVE LOGITS

 כשה

0.39

 מל

0.39

 שה

0.39

ømme

0.38

 absolue

0.37

 Mexicano

0.37

 absoluto

0.36

 tasar

0.36

 ugyan

0.36

 gustaba

0.35

Activations Density 0.055%

braces or brackets used in programming syntax

return value or property

No Comments

No Known Activations