INDEX

Explanations

phrases indicating recurring events or themes

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

usual, again, always

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

 stort

-0.31

 initial

-0.31

허

-0.29

 configureStore

-0.29

 gradual

-0.28

 sincer

-0.27

sur

-0.27

diag

-0.26

 Initial

-0.26

Initial

-0.26

POSITIVE LOGITS

今回も

0.79

usual

0.70

 Taktlose

0.68

今年も

0.66

httphttps

0.63

 yine

0.61

NewGuid

0.60

again

0.60

ⓧ

0.59

 jadx

0.59

Activations Density 0.096%

phrases indicating recurring events or themes

usual, again, always

No Comments

No Known Activations