INDEX

Explanations

group identifiers and data fields

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

variable or identifier names containing "group" followed by an underscore or dot separator in code.

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

The neuron activates on numeric literal tokens (e.g. numbers used as IDs or other constants).

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

that

-2.14

we

-2.13

--

-2.02

 debemos

-1.87

但是在

-1.67

ujeme

-1.65

IMENTO

-1.64

 precisamos

-1.62

 podríamos

-1.60

摃

-1.60

POSITIVE LOGITS

și

1.91

 ihre

1.84

寕

1.77

 और

1.75

けれ

1.75

 rendus

1.74

尐

1.72

賡

1.71

Зно

1.65

鼱

1.65

Activations Density 0.083%

group identifiers and data fields

variable or identifier names containing "group" followed by an underscore or dot separator in code.

The neuron activates on numeric literal tokens (e.g. numbers used as IDs or other constants).

No Comments

No Known Activations

group identifiers and data fields

variable or identifier names containing "group" followed by an underscore or dot separator in code.

The neuron activates on numeric literal tokens (e.g. numbers used as IDs or other constants).

No Comments

No Known Activations