INDEX

Explanations

instances of phrases indicating unexpected outcomes or revelations

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

"Turns out" or similar expressions

np_acts-logits-general · gemini-2.0-flash

turns out

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.16.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

ślę

-0.55

 olsun

-0.53

cesso

-0.52

 känns

-0.50

 Entrée

-0.50

маю

-0.49

harapkan

-0.48

Portály

-0.47

JMenu

-0.46

ERTE

-0.46

POSITIVE LOGITS

原来

0.94

原來

0.91

 &___

0.88

 actually

0.87

 ternyata

0.86

 AssemblyCulture

0.76

なんと

0.76

 bleek

0.75

実は

0.73

 оказыва

0.72

Activations Density 0.194%

instances of phrases indicating unexpected outcomes or revelations

"Turns out" or similar expressions

turns out

No Comments

No Known Activations