INDEX

Explanations

proper nouns, specifically names of people and places

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

first names and 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-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

aarrggbb

-0.74

 Administrativna

-0.56

delwed

-0.48

ſelf

-0.46

featureID

-0.45

 MaterialApp

-0.45

 ſon

-0.45

WriteTagHelper

-0.44

 étoit

-0.43

 Signalez

-0.42

POSITIVE LOGITS

 office

0.44

 officer

0.41

 president

0.41

CEO

0.41

Архівовано

0.39

SpringBootTest

0.38

 Office

0.37

awtextra

0.36

 Azevedo

0.36

autorest

0.36

Activations Density 0.011%

proper nouns, specifically names of people and places

first names and surnames

No Comments

No Known Activations