INDEX

Explanations

terms related to regulations or guidelines set by authority figures

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

rules or requirements set

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

 Passengers

-0.47

＾

-0.47

AnimationsModule

-0.44

 exportaciones

-0.43

 namespaces

-0.43

 mergeFrom

-0.42

 asynchronously

-0.41

 Sommers

-0.41

 popcorn

-0.41

playSound

-0.40

POSITIVE LOGITS

 prescribed

0.73

 stipulated

0.70

 prescribe

0.67

 predetermined

0.66

 festgelegt

0.65

 stip

0.65

 menetapkan

0.62

 belir

0.59

 festge

0.59

 mandated

0.58

Activations Density 0.021%

terms related to regulations or guidelines set by authority figures

rules or requirements set

No Comments

No Known Activations