INDEX

Explanations

programming-related terms, particularly those related to definitions and declarations in code

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

code keywords and constructs

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

 kasarigan

-0.79

ंदीखरीदारी

-0.76

oredCriteria

-0.69

 pinulongan

-0.65

 мәкал

-0.62

 مشين

-0.61

 FetchType

-0.60

 queſta

-0.60

 Италијани

-0.60

RTLD

-0.59

POSITIVE LOGITS

 panoramique

0.46

 coordinación

0.42

 ateş

0.41

 dicha

0.38

 laboral

0.38

 Verteilung

0.38

 intellectuelle

0.37

 souverain

0.36

 tumba

0.36

 palsu

0.36

Activations Density 0.274%

programming-related terms, particularly those related to definitions and declarations in code

code keywords and constructs

No Comments

No Known Activations