INDEX

Explanations

unit followed by separator or subsequent term

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

units of physical measurement (particularly SI units like Newtons, Pascals, Joules, and Tesla).

oai_token-act-pair · claude-4-5-haiku Triggered by @jyhe0408

The neuron activates on tokens that are unit symbols (e.g. N, J, W, Pa, etc.), i.e. abbreviations denoting physical measurement units.

oai_token-act-pair · o4-mini Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

沕

-1.83

CurtirCurtir

-1.80

ents

-1.70

 aantal

-1.62

珝

-1.59

 gewel

-1.57

 cordón

-1.56

atecas

-1.56

臵

-1.55

網址

-1.54

POSITIVE LOGITS

 beginnings

1.66

 after

1.61

毕竟

1.55

1.52

 eterno

1.41

خابات

1.39

agissait

1.35

putra

1.30

 whack

1.30

 القدم

1.29

Activations Density 0.025%

unit followed by separator or subsequent term

units of physical measurement (particularly SI units like Newtons, Pascals, Joules, and Tesla).

The neuron activates on tokens that are unit symbols (e.g. N, J, W, Pa, etc.), i.e. abbreviations denoting physical measurement units.

No Comments

No Known Activations

unit followed by separator or subsequent term

units of physical measurement (particularly SI units like Newtons, Pascals, Joules, and Tesla).

The neuron activates on tokens that are unit symbols (e.g. N, J, W, Pa, etc.), i.e. abbreviations denoting physical measurement units.

No Comments

No Known Activations