INDEX

Explanations

parentheses and quotes

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

line breaks or newlines in text, particularly in structured documents like lists, tables, or formatted entries.

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

This neuron responds to structured numerical tokens and their accompanying punctuation in list‐style or parenthetical contexts (e.g. years in parentheses, track times, election results).

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

-2.11

</h1>

-2.09

 whichever

-1.98

犼

-1.92

(&

-1.92

他对

-1.91

 proverbial

-1.90

 this

-1.90

 Giving

-1.88

導致

-1.88

POSITIVE LOGITS

2.45

</em>

2.23

 cuillère

2.19

鉨

2.19

these

2.05

gruppe

2.02

萏

2.00

🫓

1.99

")

1.98

érieurs

1.98

Activations Density 0.023%

parentheses and quotes

line breaks or newlines in text, particularly in structured documents like lists, tables, or formatted entries.

This neuron responds to structured numerical tokens and their accompanying punctuation in list‐style or parenthetical contexts (e.g. years in parentheses, track times, election results).

No Comments

No Known Activations

parentheses and quotes

line breaks or newlines in text, particularly in structured documents like lists, tables, or formatted entries.

This neuron responds to structured numerical tokens and their accompanying punctuation in list‐style or parenthetical contexts (e.g. years in parentheses, track times, election results).

No Comments

No Known Activations