INDEX

Explanations

code-related syntax and structures

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

declarations or code assignments

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

تقاوى

-0.71

хьтан

-0.60

 tartalomajánló

-0.58

 Roskov

-0.58

 kaarangay

-0.53

WebElementEntity

-0.52

RTLU

-0.51

casian

-0.48

 Biôgrafia

-0.48

Abitanti

-0.48

POSITIVE LOGITS

Aiheesta

0.51

 zoude

0.45

findpost

0.43

ьаж

0.42

IBOutlet

0.42

WebVitals

0.39

UnusedPrivate

0.39

 "'");

0.38

geführt

0.38

 đủ

0.38

Activations Density 0.048%

code-related syntax and structures

declarations or code assignments

No Comments

No Known Activations