"and" inside fixed field names

What this neuron does

This neuron lights up on the word "and" when it sits inside a fixed name like "earth and planetary science" or "Food and Agriculture Organization". When it was turned up, the model's writing drifted into strange number lists and repeated symbols, so the steering output doesn't show a clear theme.

Written by Claude from this neuron's activations and steering outputs. Confidence: medium. Think it is wrong? Flag it (opens a prefilled issue on GitHub).

Fires most on

The text windows where this neuron's activation was highest. Amber marks positive activation; the outlined token is the peak. Hover a token for its strength.

1. #1 peak

   series documenting colonization of the gut of a premature newborn, and one of few completely assembled community genomic datasets, said Jill Banfield, a UC Berkeley professor of earth and planetary science and of environmental science, policy and management.⏎"Sequencing of

2. #2 peak

   including Purdue University, the National Center for Atmospheric Research and the Indian Institute of Technology Delhi.⏎Dev Niyogi, a Purdue associate professor of agronomy and earth and atmospheric sciences, said modeling of a monsoon depression track can have a

3. #3 peak

   Research, Volume 1⏎Jonathan D. Jaipaul ✉️ Department of Environmental Studies. Faculty of Earth and Environmental Studies. Turkeyen Campus. Greater Georgetown, Guyana.⏎Devya Hem

4. #4 peak

   the American Society of Mechanical Engineers, ASME is a not-for-profit professional organization promoting the art, science and practice of mechanical and multidisciplinary engineering and allied sciences. ASME develops codes and standards that enhance

5. #5 peak

   worldwide including Filipino children and lactating mothers. The 2019 Expanded National Nutrition Survey (ENNS) of the Department of Science and technology-Food and Nutrition Research Institute (DOST-FNRI) shows that 1 in

6. #6 peak

   , India, focused on the empowerment of the rural and urban poor. Between 1976 and 2001, she worked with the Food and Agriculture Organization (FAO) of the United Nations to identify and support innovative NGO

Steering

Push this neuron up or down while the model writes, and watch what changes. The outputs were generated ahead of time, so nothing runs in your browser.

Most negative on

Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.

1. #1 peak

   maximum of $22.5 million and a minimum of $13.7 million a year through 2

2. #2 peak

   structures of data that can aid learning.⏎- The $M$ features of user linguistic response can be saved within a user session as a time series of vectors $\sigma(m, n, t)$.⏎- Those series can be collected across $U

3. #3 peak

   .⏎You could, of course, reach for a calculator here, but I want to show you how to do it without. You can work out $2.4 \times 100$ by moving the dot two spaces to the right

4. #4 peak

   ), the nickel ($0.05), the dime ($0.10), the quarter ($0

Which neurons in earlier layers feed this one, and which later ones it feeds. Amber links excite, blue links suppress; longer bars are stronger.

Activation histogram

How often this neuron fires at each strength. Almost always it does close to nothing (the tall bars in the middle); the snippets above are the rare moments it fired hard, amber for positive and blue for negative.

Logit lens

The words this neuron nudges the model toward (amber) or away from (blue) when it fires. A hint, not a label: many neurons work through later layers instead.