INDEX

Explanations

references to specific entities, particularly in a legal or regulatory context

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

citations and technical terms

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

 leña

-0.37

 знаешь

-0.36

 internetowa

-0.36

 laikā

-0.35

 rahasia

-0.35

 eivät

-0.35

 clínicas

-0.34

过程中

-0.34

 réussi

-0.33

 dopiero

-0.33

POSITIVE LOGITS

 surla

0.67

richTextPanel

0.66

 betweenstory

0.63

dafx

0.60

SourceChecksum

0.59

 ModelExpression

0.59

IsMutable

0.59

 BorderSide

0.58

tagHelperRunner

0.58

Geplaatst

0.58

Activations Density 0.155%

references to specific entities, particularly in a legal or regulatory context

citations and technical terms

No Comments

No Known Activations