INDEX

Explanations

exit statement

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

the word "exit" in programming code, particularly in scripts and applications.

oai_token-act-pair · claude-4-5-sonnet Triggered by @bfgcn7frf5

The neuron activates on commands or function calls that terminate execution (e.g. “exit” statements with a status code).

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

越是

-1.69

附近的

-1.68

氿

-1.66

unjung

-1.60

formaz

-1.59

-1.58

fibra

-1.58

鹋

-1.57

心中的

-1.57

人

-1.56

POSITIVE LOGITS

er

2.33

2.02

exit

1.92

ed

1.86

else

1.79

1.73

يًا

1.72

ers

1.64

1.61

鮪

1.59

Activations Density 0.002%

exit statement

the word "exit" in programming code, particularly in scripts and applications.

The neuron activates on commands or function calls that terminate execution (e.g. “exit” statements with a status code).

No Comments

No Known Activations

exit statement

the word "exit" in programming code, particularly in scripts and applications.

The neuron activates on commands or function calls that terminate execution (e.g. “exit” statements with a status code).

No Comments

No Known Activations