PLOT localizes causal variables in neural networks by fitting optimal transport couplings between abstract and neural intervention effect geometries, enabling fast handles or guided search.
Maheep Chaudhary and Atticus Geiger
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3verdicts
UNVERDICTED 3representative citing papers
AGCLR extends CoCoNuT with a gated concept stream for persistent memory to fix fact loss in latent reasoning, yielding improvements on reasoning benchmarks as depth increases.
RL preserves a larger fraction of base model circuits than SFT during fine-tuning on scientific QA, per a new head-level differential circuit vulnerability metric, at the cost of slower adaptation.
citing papers explorer
-
PLOT: Progressive Localization via Optimal Transport in Neural Causal Abstraction
PLOT localizes causal variables in neural networks by fitting optimal transport couplings between abstract and neural intervention effect geometries, enabling fast handles or guided search.
-
Why Limit the Residual Stream to Layers and Not Tokens? Persistent Memory for Continuous Latent Reasoning
AGCLR extends CoCoNuT with a gated concept stream for persistent memory to fix fact loss in latent reasoning, yielding improvements on reasoning benchmarks as depth increases.
-
Mechanistic origins of catastrophic forgetting: why RL preserves circuits better than SFT?
RL preserves a larger fraction of base model circuits than SFT during fine-tuning on scientific QA, per a new head-level differential circuit vulnerability metric, at the cost of slower adaptation.