INDEX

Explanations

dates, series, prices

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

explicit references to dates and times—calendar months, years, timestamps, and other recency/real-time context markers within responses.

oai_token-act-pair · gpt-5

Triggered by @yooniel31

2023–2024 dates

np_max-act · gpt-5

Triggered by @yooniel31

This neuron detects mentions of the model metadata token (e.g. “model”) and associated numeric or timestamp values.

oai_token-act-pair · o4-mini

Triggered by @yooniel31

New Auto-Interp

Configuration

google/gemma-scope-2-27b-it/resid_post/layer_16_width_262k_l0_medium

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Negative Logits

هر

0.31

 яр

0.27

ا

0.25

 ২০১৫

0.24

 फक्त

0.24

ارية

0.23

非常

0.23

 über

0.22

经过

0.22

 сразу

0.21

POSITIVE LOGITS

 vaccines

0.29

 lockdowns

0.27

 Biden

0.27

 Vaccination

0.26

 Cora

0.26

COR

0.26

 OpenAI

0.26

 coronavirus

0.25

 ChatGPT

0.25

openai

0.25

Activations Density 2.510%

dates, series, prices

explicit references to dates and times—calendar months, years, timestamps, and other recency/real-time context markers within responses.

2023–2024 dates

This neuron detects mentions of the model metadata token (e.g. “model”) and associated numeric or timestamp values.

No Known Activations

dates, series, prices

explicit references to dates and times—calendar months, years, timestamps, and other recency/real-time context markers within responses.

2023–2024 dates

This neuron detects mentions of the model metadata token (e.g. “model”) and associated numeric or timestamp values.

No Known Activations