INDEX

Explanations

ICP and London

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

This neuron is triggered by capitalized tokens—particularly acronyms, initialisms, and proper nouns.

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

all-caps technical acronyms and abbreviations, especially hyphenated instrument or system names from engineering and scientific contexts.

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

abbreviations and acronyms in technical or specialized contexts, particularly "ICP" (Inductively Coupled Plasma) and location/organization names containing "ati" or similar suffixes.

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

-3.22

 thinks

-2.55

There

-2.39

祈祷

-2.13

当你

-2.06

 seems

-2.05

Just

-2.05

龵

-2.00

 says

-1.96

 supposedly

-1.96

POSITIVE LOGITS

缡

2.81

沨

2.64

 acclaimed

2.59



2.38

 ซึ่ง

2.36

饹

2.33

樒

2.27

摇了

2.25

恹

2.23

읖