INDEX

Explanations

proper nouns, specifically names

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

people's names

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

httphttps

-0.60

 المعيارى

-0.53

CHOR

-0.43

◀

-0.43

ⓧ

-0.42

hpp

-0.42

adaptiveStyles

-0.41

@"/

-0.41

才有

-0.41

 nakalista

-0.41

POSITIVE LOGITS

rungsseite

0.62

 enfans

0.53

 person

0.47

 industriels

0.45

 Grecs

0.44

 AssemblyVersion

0.44

 RSSSF

0.44

 lendemain

0.43

ftagPool

0.43

 persons

0.42

Activations Density 0.051%

proper nouns, specifically names

people's names

No Comments

No Known Activations