INDEX

Explanations

programming-related terms and code constructs

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

system services and data types

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

onViewCreated

-0.71

webElementXpaths

-0.60

 ویکی‌پدی

-0.54

DockStyle

-0.53

SourceChecksum

-0.52

 Audiodateien

-0.50

ManagedBean

-0.50

'{@

-0.49

RefNanny

-0.47

/**

-0.47

POSITIVE LOGITS

:✨

0.42

 jsPsych

0.41

 InputDecoration

0.37

 يتيمه

0.36

Controllo

0.36

 lapin

0.36

LIBRA

0.35

 vectorielles

0.34

PR

0.34

 skills

0.34

Activations Density 0.017%

programming-related terms and code constructs

system services and data types

No Comments

No Known Activations