INDEX

Explanations

instances of the keyword "protected" used in code or programming contexts

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

protected keyword in code

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

 betweenstory

-0.77

 pleaſure

-0.70

ſelf

-0.69

 houſe

-0.68

 ſtate

-0.68

AddTagHelper

-0.68

 itſelf

-0.66

 Houſe

-0.65

 faſt

-0.65

 ſtre

-0.65

POSITIVE LOGITS

protected

0.79

private

0.64

0.53

0.52

0.50

 private

0.49

public

0.48

 partial

0.48

 brief

0.47

Activations Density 0.010%

instances of the keyword "protected" used in code or programming contexts

protected keyword in code

No Comments

No Known Activations