INDEX

Explanations

any followed by a word

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

The neuron fires on occurrences of the word “any” (and the tokens immediately following it).

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

indefinite determiners signaling non-specific quantity or selection, especially in conditional/request constructions (e.g., “if … [indefinite determiner] …” and “[…] kind of …”).

oai_token-act-pair · gpt-5 Triggered by @jyhe0408

the phrase "if you have any" followed by a noun, typically in customer service or informational contexts.

oai_token-act-pair · claude-4-5-sonnet 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

 where

-1.66

 while

-1.46

こども

-1.39

how

-1.39

 produrre

-1.35

 spiega

-1.33

因为它

-1.33

it

-1.33

 until

-1.31

 although

-1.28

POSITIVE LOGITS

indonesia

1.37

1.32

 innen

1.32

liefert

1.30

各種

1.27

 evergreen

1.27

ett

1.25

ttää

1.20

 после

1.19

 meiner

1.16

Activations Density 0.022%

any followed by a word

The neuron fires on occurrences of the word “any” (and the tokens immediately following it).

indefinite determiners signaling non-specific quantity or selection, especially in conditional/request constructions (e.g., “if … [indefinite determiner] …” and “[…] kind of …”).

the phrase "if you have any" followed by a noun, typically in customer service or informational contexts.

No Comments

No Known Activations