INDEX

Explanations

programming constructs and elements related to observable objects and data structures

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

body or mass

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

AsUp

-0.68

 autorytatywna

-0.68

setPointSize

-0.65

ArgsConstructor

-0.64

StructEnd

-0.63

SharedDtor

-0.61

AndEndTag

-0.60

 protoimpl

-0.60

addCriterion

-0.59

InjectAttribute

-0.59

POSITIVE LOGITS

body

0.53

 body

0.47

Bodies

0.47

 Bodies

0.44

Body

0.42

BODY

0.40

 bodies

0.40

 Body

0.39

bodies

0.39

身

0.39

Activations Density 0.000%

programming constructs and elements related to observable objects and data structures

body or mass

No Comments

No Known Activations