INDEX

Explanations

references to actions, user interactions, and decisions in a contextual framework

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

instructions and technical terms

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

 intptr

-0.52

ItemBackground

-0.42

Tembelea

-0.39

LEncoder

-0.37

qrstuvwxyz

-0.36

PickerController

-0.35

addContainerGap

-0.35

>//

-0.35

 diecast

-0.34

writeFile

-0.34

POSITIVE LOGITS

 للاسماء

0.57

 kasarigan

0.50

 فريبيس

0.48

fromnode

0.47

 nakalista

0.46

UnusedPrivate

0.46

aarrggbb

0.44

ruptedException

0.43

septic

0.41

 Geset

0.40

Activations Density 0.082%

references to actions, user interactions, and decisions in a contextual framework

instructions and technical terms

No Comments

No Known Activations