INDEX

Explanations

verbs indicating effort or activity in various tasks or processes

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

ongoing actions lately

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

 comodidad

-0.47

 hijas

-0.47

 nanti

-0.43

 nantinya

-0.43

 pozdě

-0.43

 Widerspruch

-0.42

 circunstancias

-0.42

almaz

-0.42

 decoración

-0.41

ندران

-0.41

POSITIVE LOGITS

 lately

0.63

depuis

0.59

 since

0.56

以来

0.54

 depuis

0.54

以來

0.52

 sinds

0.52

 Lately

0.51

 seit

0.51

Depuis

0.50

Activations Density 0.022%

verbs indicating effort or activity in various tasks or processes

ongoing actions lately

No Comments

No Known Activations