INDEX

Explanations

species names

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

place names from non-Western languages.

oai_token-act-pair · claude-3-7-sonnet-20250219

Triggered by @neilrathi

The neuron strongly activates on uncommon, content‐bearing single words—especially capitalized or specialized proper nouns (e.g. species names and other rare terms).

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

Negative Logits

 resistentes

-2.13

and

-2.02

-1.97

 eficientes

-1.86

 ambos

-1.74

 complicado

-1.59

亮点

-1.57

 competitivos

-1.55

</h1>

-1.55

 seguras

-1.54

POSITIVE LOGITS

with

1.98

 Notwithstanding

1.86

稍微

1.83

 Houſe

1.82

を行い

1.77

 Owing

1.75

儛

1.70

 définiti

1.68

酲

1.67

 CALIFORNI

1.65

Activations Density 0.001%

species names

place names from non-Western languages.

The neuron strongly activates on uncommon, content‐bearing single words—especially capitalized or specialized proper nouns (e.g. species names and other rare terms).

No Known Activations

species names

place names from non-Western languages.

The neuron strongly activates on uncommon, content‐bearing single words—especially capitalized or specialized proper nouns (e.g. species names and other rare terms).

No Known Activations