INDEX

Explanations

location or computer hardware

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

phrases indicating location or position within a container or space.

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

The neuron strongly activates on deictic location words and phrases (e.g. “in,” “on,” “there,” “your desktop,” “in your build,” etc.) that point to where something is or should go.

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

(\%)

-1.33

tă

-1.30

 oggi

-1.27



-1.21

Yup

-1.20

 különböző

-1.20

鐽

-1.19

 börjar

-1.17

🐑

-1.16

different

-1.15

POSITIVE LOGITS

it

1.86

 there

1.74

we

1.34

 where

1.34

in

1.32

 Most

1.20

are

1.18

so

1.18

 With

1.15

Activations Density 0.428%

location or computer hardware

phrases indicating location or position within a container or space.

The neuron strongly activates on deictic location words and phrases (e.g. “in,” “on,” “there,” “your desktop,” “in your build,” etc.) that point to where something is or should go.

No Comments

No Known Activations

location or computer hardware

phrases indicating location or position within a container or space.

The neuron strongly activates on deictic location words and phrases (e.g. “in,” “on,” “there,” “your desktop,” “in your build,” etc.) that point to where something is or should go.

No Comments

No Known Activations