'fixed' as a finance or cost term
A neuronOne of thousands of tiny dials inside each layer. It reads the text so far and produces a number, its activation, that nudges what the model writes next. in layerOne step in the stack the text passes through. Qwen3-0.6B has 28 layers, each with 3,072 neurons. 4 of Qwen3-0.6B. Its strongest snippets are very consistent: they mostly peak on the same token. It is featuredA neuron picked for the full treatment: steering outputs and a written explanation. About 3,000 of the 86,000 neurons are featured., so it has steering outputs.
What this neuron does
This neuron fires on the word 'fixed' in phrases like 'fixed asset', 'fixed amount', 'fixed-income' and 'fixed price', especially in finance and insurance text. Turning it up did not bring out finance-related writing in the sample outputs, so the link is only seen in the examples.
Written by Claude from this neuron's activations and steering outputs. Confidence: high. Think it is wrong? Flag it (opens a prefilled issue on GitHub).
Fires most on
The text windows where this neuron's activationThe number a neuron produces for a token. Large positive means it fired hard; negative means it pushed the other way. was highest. Amber marks positive activation; the outlined tokenA piece of text the model reads at once: often a word, sometimes part of a word or a punctuation mark. is the peak. Hover a token for its strength.
-
value function for a fixed policy, in a simulated domain.⏎the application examples in programming are very good and
-
to interpret this extremely useful financial measurement.⏎Financial ratios such as the fixed asset turnover help financial analysts, management, and investors alike to make critical decisions whether
-
calculating the current bond price with our bond valuation calculator, let's take some time to talk about what a bond is. A bond is one of the most prevalent fixed-income securities. When an entity issues bonds, it is considered as acquiring funding from
-
company and, usually, to the assets of the company in the event of its dissolution. Dividends are usually set at a fixed annual rate that must be paid before dividends are distributed to the common stockholders.
-
to maintain coverage.⏎b. Deductible: The predetermined amount that individuals must pay out-of-pocket before insurance coverage kicks in.⏎c. Copayment: A fixed amount individuals pay for a specific healthcare service, such as a doctor’s visit or
-
products and services at similar prices throughout the supply chain.⏎How is this different from the fixed cost? In the case of fixed price, you have to factor in costs of production before you come up with your
SteeringAdding a push along a neuron while the model writes, to see how its output changes. Turning it up or down.
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.
Here is something I have been thinking about lately: can I help with this question? I can't help with this question.] Okay, this is a bit of a stretch. I need to think that I can't help with this question.] Okay, I need to tell you to
Most negative on
Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.
-
children learn and develop. There’s no reason why children with autism who use visual supports are any different. So why not bring play and fun games for children with autism into your
-
reasonable.⏎But, of course, the reasonable nature of Utterson’s deductions proves precisely their downfall. Utterson remains so adamantly rational and sensible that he never once admits the possibility of a supernatural explanation. He is the embodiment of the Victorian mind
-
existence, blazing a trail of death and destruction that prompted mobs of angry villagers to take up torches and pitchforks against him on the silver screen. Never once during his rampage, however, did the monster question the honesty of his ultimate creator
-
.⏎This disobedient crowd, momentarily growing as others pushed forward from the main body of people who stayed respectfully at a distance, failed to heed Moses. As soon as Moses reached them, he again shouted his warning.⏎“Move back at once all
Connections
Show on mapWhich neuronsOne of thousands of tiny dials inside each layer. It reads the text so far and produces a number, its activation, that nudges what the model writes next. in earlier layersOne step in the stack the text passes through. Qwen3-0.6B has 28 layers, each with 3,072 neurons. feed this one, and which later ones it feeds. Amber links excite, blue links suppress; longer bars are stronger.
Fed by earlier layers
- L2:3033 suppresses
- L3:2144 suppresses
- L3:1281 weakly suppresses
- L1:1401 weakly excites
- L2:745 weakly excites
- L2:597 weakly excites
- L1:515 weakly suppresses 'about' in 'it's not about' phrasing
- L3:2736 weakly excites
- L1:1287 weakly excites
- L3:115 weakly suppresses
- L2:2513 weakly suppresses
- L2:1708 weakly suppresses
- L2:2799 weakly excites
- L1:1386 weakly suppresses
- L3:456 weakly suppresses
- L2:2900 weakly excites
- L2:115 weakly excites
- L3:1592 weakly excites
- L3:464 weakly excites
- L0:1345 weakly excites
- L0:2320 weakly excites
- L0:1277 weakly suppresses
- L3:2003 weakly suppresses
- L2:2982 weakly suppresses
Feeds later layers
- L8:1461 excites
- L11:1430 excites
- L5:1260 weakly excites
- L23:1674 weakly excites
- L26:2879 weakly excites
- L12:2525 weakly excites
- L16:2089 weakly suppresses
- L5:946 weakly suppresses
- L15:842 weakly suppresses 'not' in 'if not most/all' phrases
- L19:2846 weakly suppresses
- L8:1167 weakly excites
- L5:1554 weakly suppresses
- L11:2468 weakly excites
- L21:225 weakly excites
- L5:1947 weakly suppresses
- L12:54 weakly excites
- L15:1114 weakly excites
- L16:921 weakly suppresses
- L16:2848 weakly excites
- L20:953 weakly excites
- L17:2463 weakly excites
- L25:694 weakly suppresses
- L11:812 weakly suppresses
- L6:1024 weakly excites
Activation histogramA chart of how often a neuron takes each activation value.
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 lensReading which output tokens a neuron pushes toward or away from, straight off its output weights, before the later layers have their say.
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 layersOne step in the stack the text passes through. Qwen3-0.6B has 28 layers, each with 3,072 neurons. instead.
Promotes
- 工委
- 탕
- Scanner
- aneously
- раг
- _fixed
- .fixed
- licensee
- -fixed
- ArrayOf
Suppresses
- FG
- velt
- зем
- ifique
- ém
- SV
- .geom
- 括
- �
- thế