Pith. sign in

REVIEW 13 cited by

Learning Successor States and Goal-Dependent Values: A Mathematical Viewpoint

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2101.07123 v1 pith:E6JC2KJG submitted 2021-01-18 cs.LG

Learning Successor States and Goal-Dependent Values: A Mathematical Viewpoint

classification cs.LG
keywords stateslearningsuccessorfunctionoperatorstatevaluebellman
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

In reinforcement learning, temporal difference-based algorithms can be sample-inefficient: for instance, with sparse rewards, no learning occurs until a reward is observed. This can be remedied by learning richer objects, such as a model of the environment, or successor states. Successor states model the expected future state occupancy from any given state for a given policy and are related to goal-dependent value functions, which learn how to reach arbitrary states. We formally derive the temporal difference algorithm for successor state and goal-dependent value function learning, either for discrete or for continuous environments with function approximation. Especially, we provide finite-variance estimators even in continuous environments, where the reward for exactly reaching a goal state becomes infinitely sparse. Successor states satisfy more than just the Bellman equation: a backward Bellman operator and a Bellman-Newton (BN) operator encode path compositionality in the environment. The BN operator is akin to second-order gradient descent methods and provides the true update of the value function when acquiring more observations, with explicit tabular bounds. In the tabular case and with infinitesimal learning rates, mixing the usual and backward Bellman operators provably improves eigenvalues for asymptotic convergence, and the asymptotic convergence of the BN operator is provably better than TD, with a rate independent from the environment. However, the BN method is more complex and less robust to sampling noise. Finally, a forward-backward (FB) finite-rank parameterization of successor states enjoys reduced variance and improved samplability, provides a direct model of the value function, has fully understood fixed points corresponding to long-range dependencies, approximates the BN method, and provides two canonical representations of states as a byproduct.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 13 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Switching Successor Measures for Hierarchical Zero-shot Reinforcement Learning

    cs.LG 2026-05 unverdicted novelty 7.0

    Switching successor measures extend classical successor measures to enable hierarchical zero-shot RL via the FB π-Switch algorithm that extracts subgoal-selection and control policies from forward-backward representations.

  2. Understanding Human Actions through the Lens of Executable Models

    cs.AI 2026-04 unverdicted novelty 7.0

    EXACT is a new DSL for human motions as executable reward-generating programs, enabling compositional neuro-symbolic models that improve data efficiency and capture intuitive action relationships over monolithic approaches.

  3. SVL: Goal-Conditioned Reinforcement Learning as Survival Learning

    cs.LG 2026-04 unverdicted novelty 7.0

    Survival value learning expresses the goal-conditioned value function as a discounted sum of survival probabilities and estimates it with maximum-likelihood hazard models on censored data, matching or exceeding TD bas...

  4. Exploration and Online Transfer with Behavioral Foundation Models

    cs.AI 2026-06 unverdicted novelty 6.0

    Proposes framing online zero-shot RL transfer as a bandit problem solved by BFMs, deriving eigenvalue minimization of an uncertainty matrix for exploration under linear reward approximation, validated on a simple environment.

  5. Exploration and Online Transfer with Behavioral Foundation Models

    cs.AI 2026-06 unverdicted novelty 6.0

    Frames online zero-shot transfer with BFMs as a bandit problem and derives an eigenvalue-minimization exploration strategy under linear reward approximation.

  6. Plan, Don't Pose: Long Composite Motion Generation with Text-Aligned BFM

    cs.LG 2026-05 unverdicted novelty 6.0

    Text2BFM aligns language with a frozen BFM via a text-aligned variational behavioral bottleneck to generate long motions by decoding latents into policy actions.

  7. Offline Reinforcement Learning with Universal Horizon Models

    cs.LG 2026-05 unverdicted novelty 6.0

    Universal horizon models extend geometric horizon models to arbitrary horizons and apply winsorized distributions for stable offline RL value learning, outperforming baselines on 100 OGBench tasks.

  8. QHyer: Q-conditioned Hybrid Attention-mamba Transformer for Offline Goal-conditioned RL

    cs.LG 2026-05 unverdicted novelty 6.0

    QHyer replaces return-to-go with a state-conditioned Q-estimator and adds a gated hybrid attention-mamba backbone to achieve state-of-the-art performance in offline goal-conditioned RL on both Markovian and non-Markov...

  9. QHyer: Q-conditioned Hybrid Attention-mamba Transformer for Offline Goal-conditioned RL

    cs.LG 2026-05 unverdicted novelty 6.0

    QHyer achieves state-of-the-art results in offline goal-conditioned RL by replacing return-to-go with a state-conditioned Q-estimator and introducing a gated hybrid attention-mamba backbone for content-adaptive histor...

  10. When Dynamics Shift, Robust Task Inference Wins: Offline Imitation Learning with Behavior Foundation Models Revisited

    cs.LG 2026-05 unverdicted novelty 5.0

    Robust minimax task inference in BFMs achieves dynamics-shift robustness from nominal offline data alone and outperforms standard baselines.

  11. Self-Predictive Representations for Combinatorial Generalization in Behavioral Cloning

    cs.LG 2025-06 unverdicted novelty 5.0

    BYOL-γ uses self-predictive representations to approximate successor representations, improving zero-shot combinatorial generalization in goal-conditioned behavioral cloning.

  12. Intention-Conditioned Flow Occupancy Models

    cs.LG 2025-06 unverdicted novelty 5.0

    InFOM applies flow matching to model intention-conditioned occupancy measures for RL pre-training, reporting 1.8x median return gains and 36% higher success rates on benchmarks.

  13. Spectral Alignment in Forward-Backward Representations via Temporal Abstraction

    cs.LG 2026-03 unverdicted novelty 4.0

    Temporal abstraction functions as a low-pass filter on transition dynamics to lower the effective rank of successor representations while bounding value function error in forward-backward learning.