INDEX

Explanations

phrases indicating structure and interactions within a composition or system

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

affected by or with

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

parsedMessage

-0.45

Parcelize

-0.42



-0.35

 &___

-0.34

}*/

-0.33

HasForeignKey

-0.32

 among

-0.32

 Both

-0.31

 presenti

-0.31

 trajets

-0.30

POSITIVE LOGITS

 EconPapers

0.67

Selfer

0.53

oredCriteria

0.49

ỏa

0.49

 confronti

0.48

sizeCache

0.48

ConstraintMaker

0.48

SharedDtor

0.47

ßig

0.46

 CURLOPT

0.46

Activations Density 0.123%

phrases indicating structure and interactions within a composition or system

affected by or with

No Comments

No Known Activations