INDEX

Explanations

specific proper nouns and identifiers related to organizations and entities

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

C reactive protein

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

BeginContext

-0.45

invokeLater

-0.42

 snippetHide

-0.41

 personnalisés

-0.38

addCriterion

-0.37

antMatchers

-0.37

 disambiguazione

-0.35

 BoxDecoration

-0.35

PropertyChanging

-0.35

 تضيفلها

-0.35

POSITIVE LOGITS

 queſto

0.54

DSA

0.51

 visitor

0.51

acad

0.50

 announce

0.50

Visitor

0.50

Pathway

0.50

visitor

0.50

 Visitor

0.50

 enqu

0.50

Activations Density 0.022%

specific proper nouns and identifiers related to organizations and entities

C reactive protein

No Comments

No Known Activations