INDEX

Explanations

This neuron detects enthusiastic product-promotion words highlighting ease, creativity, and playful features.

oai_token-act-pair · o4-mini

Triggered by @jyhe0408

words related to specialized topics, fields, or domains in descriptive or informational text.

oai_token-act-pair · claude-4-5-sonnet

Triggered by @jyhe0408

formal, abstract institutional language—nominalizations and conceptual terms typical of academic, business, or promotional/event contexts

oai_token-act-pair · gpt-5

Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

и

2.98

และ

2.98

and

2.97

 અને

2.97

 και

2.91

 और

2.89

và

2.88

 आणि

2.77

ও

2.73

 ਅਤੇ

2.69

POSITIVE LOGITS

 alike

1.68

 sowie

1.42

 यांचे

1.39

 ஆகியவை

1.35

 ஆகியவற்ற

1.30

 यांना

1.29

 sebagainya

1.28

 serta

1.27

 উভয়

1.25

 यांनी

1.24

Activations Density 1.708%

This neuron detects enthusiastic product-promotion words highlighting ease, creativity, and playful features.

words related to specialized topics, fields, or domains in descriptive or informational text.

formal, abstract institutional language—nominalizations and conceptual terms typical of academic, business, or promotional/event contexts

No Known Activations

This neuron detects enthusiastic product-promotion words highlighting ease, creativity, and playful features.

words related to specialized topics, fields, or domains in descriptive or informational text.

formal, abstract institutional language—nominalizations and conceptual terms typical of academic, business, or promotional/event contexts

No Known Activations