INDEX

Explanations

names, particularly family names and lineage connections

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

Surnames

np_acts-logits-general · gemini-2.0-flash

surnames

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 balancing

-0.43

 hard

-0.43

 Balancing

-0.42

 dual

-0.40

 inverted

-0.38

 difficult

-0.37

Balancing

-0.37

 reverse

-0.36

 pain

-0.36

 leher

-0.36

POSITIVE LOGITS

 ſch

0.62

 ſta

0.61

 ſte

0.61

 deſt

0.57

GEBURTSDATUM

0.56

 ſon

0.56

ſelf

0.56

colgante

0.56

berdayakan

0.55

 purpoſe

0.55

Activations Density 0.054%

names, particularly family names and lineage connections

Surnames

surnames

No Comments

No Known Activations