INDEX

Explanations

proper nouns and names associated with various entities or subjects

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

iſen, queſta, brakk

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

uxxxx

-0.46

 curiosidad

-0.42

 appunto

-0.41

 fatiga

-0.40

complexType

-0.39

 Suiza

-0.38

able

-0.38

 Belanda

-0.38

 volontaire

-0.37

 Bitte

-0.37

POSITIVE LOGITS

 queſta

0.62

 समीक्षाएं

0.57

iſen

0.56

mybatisplus

0.55



0.54

brakk

0.53

PDOException

0.52

addContainerGap

0.49

 OnTriggerEnter

0.49

ⓧ

0.49

Activations Density 0.093%

proper nouns and names associated with various entities or subjects

iſen, queſta, brakk

No Comments

No Known Activations