INDEX

Explanations

programming-related syntax elements and function calls

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

code parsing and generation

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

/**

-0.41

Indented

-0.39

 strukt

-0.38

 réservoir

-0.37

 Grüsse

-0.35

<?

-0.35

Autowired

-0.34

 Comentarios

-0.34

 estrut

-0.34

rão

-0.33

POSITIVE LOGITS

:✨

0.71

 CreateTagHelper

0.64

rungsseite

0.58

ftagPool

0.58

tagHelperRunner

0.57

ValueStyle

0.56

 Réponses

0.56

новниш

0.54

 disambiguazione

0.54

 kasarigan

0.54

Activations Density 0.003%

programming-related syntax elements and function calls

code parsing and generation

No Comments

No Known Activations