INDEX

Explanations

elements related to categories and responses in structured formats, such as quizzes or tests

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

smaller: 0.135 or -2?

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

expandindo

-0.49

 templado

-0.44

 secuencia

-0.43

 productivo

-0.42

 césped

-0.42

 conformidad

-0.42

 cerámica

-0.42

 mármol

-0.42

izr

-0.42

 dignité

-0.41

POSITIVE LOGITS

 resourceCulture

0.51

NameIn

0.46

❹

0.45

 Италијани

0.44

 rivets

0.44

FunctionFlags

0.43

⤹

0.43

rhosis

0.43

enido

0.42

UrlResolution

0.42

Activations Density 0.018%

elements related to categories and responses in structured formats, such as quizzes or tests

smaller: 0.135 or -2?

No Comments

No Known Activations