INDEX

Explanations

tokens related to programming logic and syntax in coding contexts

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

n % calculation

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

 Molière

-0.43

 bush

-0.43

Ny

-0.42

moiselle

-0.41

⎤

-0.41

 Matilda

-0.40

 ARIA

-0.40

 versa

-0.40

 laure

-0.40

angelo

-0.39

POSITIVE LOGITS

EndContext

0.57

SharedDtor

0.49

FunctionFlags

0.48

 gyhoeddwyd

0.45

utilisons

0.43

 आइटम

0.42

gnore

0.41

 bénévoles

0.41

 '\\;'

0.41

Activations Density 0.000%

tokens related to programming logic and syntax in coding contexts

n % calculation

No Comments

No Known Activations