INDEX

Explanations

references to academic citations or footnotes

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

after user input

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

awtextra

-0.69

GEBURTSDATUM

-0.68

fjspx

-0.67

Tembelea

-0.65

 ModelExpression

-0.62

 Numerade

-0.61

 OFDb

-0.60

новниш

-0.56

 للاسماء

-0.56



-0.56

POSITIVE LOGITS

 rainha

0.42

 irmã

0.42

WEBPACK

0.41

 Argentino

0.40

 civilian

0.39

 feroit

0.39

 AssemblyCompany

0.39

raszam

0.39

 współpra

0.39

 giovan

0.39

Activations Density 0.007%

references to academic citations or footnotes

after user input

No Comments

No Known Activations