If MLP neurons are monosemantic, are known entities capped by neuron count?
JessicaRumbelow · x · 2026-10-18
JessicaRumbelow kicks off an interpretability discussion: if MLP neurons are indeed monosemantic/monofunctional, should we expect the number of entities a model knows to be bounded by the number of neurons?
In follow-ups, luciaquirke proposes compositional representations (entities as co-activations of composable-concept neurons rather than dedicated entity units), and JessicaRumbelow concludes it's likely a combinatorial/composability matter: the combinatorial space of many neurons is gigantic, so you only need neurons for each composable part, not every entity or concept.
Related event: Monosemanticity debate: Does neuron count cap what a model can know?(8 posts)→
More from Research
- IFM open-sources RL360: 8,192 concurrent sandboxes and ~1.3M rollouts for agentic RL — rupspace · 2026-10-23
- Science lifts expression of concern on 2024 AI debunking-conspiracies paper; conclusions unchanged — DG_Rand · 2026-10-23
- Embodied First Exam benchmarks coding agents on 1,090 tasks across 44 robots — ManlingLi_ · 2026-10-23
- Anima Anandkumar on neural operators: AI weather models beat supercomputers — AnimaAnandkumar · 2026-10-23
- Scaling GLM-5.3-Flash training to 1M token context, 64x over Megatron baseline — AndrewLBeam · 2026-10-23
- Experimental boolean-logic LM: microsecond forward pass on one CPU thread, but painfully slow training — cephaloform · 2026-10-23