AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Dynamic-Depth Context Tree Weighting

Joao V Messias, Shimon Whiteson

NeurIPS 2017 · Reinforcement learning › Visual and representation RL · approach: Scalable RL systems and environments · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
dynamic-depth fringe expansion (D2-CTW) instead of fixed-depth CTWcontext tree weighting on Calgary Corpus · Calgary Corpus, unbounded memoryavg log-loss and node count: log-loss ~2% higher than CTW; ~28x fewer nodesMixed
enforce node bound L=1000 on the tree (D2-CTW uses fringe expansion and pruning, CTW stops growing)CTW vs D2-CTW · Calgary Corpusavg log-loss: CTW 11.4% higher on averageImproved
dynamic expansion with significance test (D2-CTW) under 5% uniform noise plus node boundCTW under noise · Calgary Corpus with 5% noise, L=1000avg log-loss: CTW 20.0% worseImproved
dynamic depth (D2-CTW) instead of fixed K=3 CTW/PPM-Cmodel-based RL in T-Maze · T-Maze length 4average reward per episodeImproved
node bound L=1000 applied to all methodsD2-CTW vs CTW and PPM-C in cheese maze · Cheese Mazeaverage reward per stepImproved
D2-CTW with node bound L=1000 instead of CTW/PPM-Cpartially observable mountain car · P.O. Mountain Car, L=1000average reward per episodeImproved
varying likelihood ratio threshold gamma (ablation)D2-CTW · Calgary Corpus (book2, 10% noise)avg log-lossNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy