INDEX

Explanations

action-oriented statements and interactions

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

indicated, asked, looked

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

SpringBootTest

-0.60

 '\\;'

-0.51

multer

-0.50

 pymongo

-0.49

/**

-0.49

 purpoſe

-0.49

Flask

-0.47

NoSuchAlgorithm

-0.47

 gyhoeddwyd

-0.47

RouterModule

-0.47

POSITIVE LOGITS

AsUp

0.48

StoryboardSegue

0.44

참고

0.43

HtmlAttribute

0.40

PostInfinity

0.35

XmlAccessorType

0.34

pat

0.33

 оказалось

0.33

 schließlich

0.33

SourceChecksum

0.33

Activations Density 0.110%

action-oriented statements and interactions

indicated, asked, looked

No Comments

No Known Activations