middle pieces of split longer words

What this neuron does

This neuron fires on the later pieces of long words that the model had to split into several chunks, such as 'reconfiguration', 'comparability' and 'grouping'. Turning it up did not produce a clear visible change in the writing.

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

   reconfiguration region are governed by field-aligned currents, which in their turn generate auroral bright

2. #2 peak

   any reconfiguration event, there is no need to run this command interactively.⏎For more information

3. #3 peak

   and texture of materials on the surface. Areas with a redder color appear red, and areas with a grayer color appear blue. Young crater rays, arrayed radially around fresh impact craters

4. #4 peak

   certain services category definition. For such reasons, compilers should have in-depth knowledge of the methodologies and definitions used by the other data sources to ensure consistency and the comparability of the resulting merged data. Various dimensions of the different sources must be analysed by

5. #5 peak

   data points, it mainly deals with unlabelled data.⏎Some applications of unsupervised machine learning techniques include: clustering (grouping coherent segments of customers, or grouping logistic pick-up points), anomaly detection (detect

6. #6 peak

   asymmetric radial transects are fixed. We provide a new method that eliminates edge-effect bias for designs that use asymmetric radial transects with fixed orientation, but a workable solution for polygonal transects remains elusive.⏎Keywords: Boundary overlap, Inclusion

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

   al and differentiation.⏎"This research will be a big boost for stem cell scientists. The protein phosphorylation sites identified in this study are freely available to the broader research community, and researchers can use these data to study the cells in greater depth and determine

2. #2 peak

   a threat from “possible large” earthquakes, studies on active fault zones in Thailand are still minimal.⏎With all of these in mind, the Geophysics research group at Faculty of Science, Mahidol University is founded with the objective of being

3. #3 peak

   drivers of plant redistribution in mountainous areas under global change.⏎Personal links. ResearchGate | GoogleScholar | Research Group Website⏎Institute. Department of Agronomy, Food, Natural resources, Animals

4. #4 peak

   0, 1, depending on the quadratic residue properties of modulo .⏎A famous result of Gauss states that when the modulus is a prime , the two sums and have

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.