INDEX

Explanations

ability to actions

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

The neuron detects occurrences of the modal verb “can.”

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

modal expressions of ability or possibility (often an auxiliary before an infinitive, including negative forms), sometimes extending to the adjacent main verb indicating the action enabled or permitted.

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

the word "can" when expressing ability or possibility.

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

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_22/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

자인

-1.25

΄

-1.25

れました

-1.22

ølge

-1.14

׳

-1.13

 personenbez

-1.05

Насе

-1.05

主机

-1.04

ebly

-1.03

寿

-1.02

POSITIVE LOGITS

 only

1.59

by

1.16

 lahko

1.09

 easily

1.05

になりますが

1.00

no

1.00

cieux

1.00

0.98

혀

0.98

Activations Density 0.033%

ability to actions

The neuron detects occurrences of the modal verb “can.”

modal expressions of ability or possibility (often an auxiliary before an infinitive, including negative forms), sometimes extending to the adjacent main verb indicating the action enabled or permitted.

the word "can" when expressing ability or possibility.

No Comments

No Known Activations

ability to actions

The neuron detects occurrences of the modal verb “can.”

modal expressions of ability or possibility (often an auxiliary before an infinitive, including negative forms), sometimes extending to the adjacent main verb indicating the action enabled or permitted.

the word "can" when expressing ability or possibility.

No Comments

No Known Activations