INDEX

Explanations

references to victory or success in competitive contexts

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

winning or losing

np_acts-logits-general · gemini-2.0-flash

win, loss, victory, defeat

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ArgumentParser

-0.69

]")]

-0.63

czaj

-0.62

HtmlAttribute

-0.59

NOSTIC

-0.57

∬

-0.54

ು

-0.48

｀)

-0.46

}))

-0.46

enez

-0.46

POSITIVE LOGITS

win

2.56

 winning

2.52

 victory

2.51

 wins

2.41

 winner

2.20

 defeat

2.12

Winning

2.12

 winners

2.11

 Winning

2.10

 victories

2.07

Activations Density 0.376%

references to victory or success in competitive contexts

winning or losing

win, loss, victory, defeat

No Comments

No Known Activations