INDEX

Explanations

elements related to programming or coding structure

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

type and opcode declarations

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

 الحره

-0.79

 فريبيس

-0.64

 كومونز

-0.56

complexContent

-0.54

UIControlState

-0.53

 propOrder

-0.53

findpost

-0.51

['$

-0.50

 المعيارى

-0.49

 Pren

-0.48

POSITIVE LOGITS

 otomatig

0.43

SceneManagement

0.39

Panamoan

0.37

 ProtoMessage

0.33

 soát

0.33

ьаж

0.33

WaitGroup

0.33

 fapt

0.33

 vorg

0.33

addTab

0.32

Activations Density 0.108%

elements related to programming or coding structure

type and opcode declarations

No Comments

No Known Activations