Pith. sign in

REVIEW 3 major objections 4 minor 43 references

Meta-Inverse Reinforcement Learning for Mean Field Games via Probabilistic Context Variables

T0 review · 3 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that a probabilistic context variable placed outside a mean field game lets inverse reinforcement learning recover reward functions from heterogeneous, unlabelled demonstrations, and that the recovered reward reproduces the

desk verdict The core estimator in Eq. (8) is biased by p(m), and the theory leans on an unproved lemma; the paper is a plausible extension with nice experiments, but the central claim needs repair. read the letter →

arxiv 2509.03845 v1 pith:ZPP45KGV submitted 2025-09-04 cs.LG cs.AIcs.GT

Yang Chen , Xiao Lin , Bo Yan , Libo Zhang , Jiamou Liu , Neset Özkan Tan , Michael Witbrock This is my paper · ORCID
classification cs.LGcs.AIcs.GT
keywords meanfieldgamesinversereinforcementlearningmeta-learningprobabilisticcontextvariableslatentvariablemodelsheterogeneousagentsadversarialIRLspatialpricing
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to answer a practical question: can an observer infer reward functions from demonstrations produced by many interacting agents who are not all pursuing the same objective? It claims the answer is yes when the different objectives are structurally similar versions of one task, indexed by a discrete probabilistic context variable. The proposed method, PEMMFIRL, learns both a reward function and a context-inference model from a pool of mixed-type trajectories, using nothing but the observed demonstrations. If this is right, inverse reinforcement learning for mean field games can be applied to real-world populations such as taxi drivers with different trip preferences, without knowing each driver's type or reworking the mean field equilibrium theory. Experimental results on three simulated mean field games and a New York taxi pricing task support the claim.

What carries the argument

The load-bearing object is the probabilistic context variable $m$, attached to a family of mean field games, together with the variational context-inference model $q_\psi(m|\tau)$ and the context-conditioned empirical mean field estimator $\hat{\mu}_\psi(s|m) = E[q_\psi(m|\tau_E) \cdot \mathbf{1}\{s_t=s\}]$. This estimator transfers the unknown task type into the mean-field term while leaving each fixed-$m$ game an ordinary mean field game; mutual-information regularisation is what prevents the reward from collapsing onto $m$-independence. Gradient estimates for the reward and the inference model are derived through the adversarial MF-AIRL sampler $\pi_\theta(a|s,m)$, so the intractable partition function never has to be computed.

What would settle it

Construct a two-context mean field game with known contexts, train PEMMFIRL until convergence, then feed it held-out expert trajectories and compare the equilibrium induced by $f_\omega(s,a,\mu,\hat{m})$ with the expert equilibrium under the true reward $r(s,a,\mu,\hat{m})$; a systematic divergence, or a divergence that does not shrink as the number of demonstrations grows, would falsify the claim.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that a probabilistic context variable placed outside—rather than inside—a mean field game lets inverse reinforcement learning recover rewards from heterogeneous, unlabelled demonstrations. For each fixed context $m$, the game is an ordinary mean field game with its usual equilibrium theory; what is learned is a variational model $q_\psi(m|\tau)$ for the context and a reward function $f_\omega(s,a,\mu,m)$. The paper asserts that for a new expert trajectory $\tau_E$, the inferred context $\hat{m}$ drawn from $q_\psi$ yields a reward $f_\omega(s,a,\mu,\hat{m})$ whose induced entropy-regularised mean field Nash equilibrium matches the equilibrium of the true reward $r(s,a,\mu,\hat{m})$. The mechanism is a mutual

Load-bearing premise

The load-bearing assumption is that the learned context-inference model becomes exactly the true posterior probability of a context given a trajectory; if it is only approximate, the context-conditioned mean field estimate is biased and the claim that the inferred reward reproduces the expert equilibrium no longer follows.

Editorial extensions

If this is right

  • Rewards can be inferred from mixed-type demonstrations without any access to the context values, the context prior, or the transition dynamics.
  • The fixed-context equilibrium properties of mean field games are inherited unchanged, because the context variable lives outside each individual game.
  • At meta-test time, a single new trajectory suffices to infer the task context and produce a reward whose induced equilibrium matches the expert's.
  • On a real New York taxi pricing task, the learned reward and policy raise drivers' average profit by roughly three percent while losing fewer than one percent of served passengers.
  • In simulated mean field games, policies recovered by this method deviate far less from the expert policy than baselines that assume a single homogeneous population.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The variational model qψ is assumed to become the true posterior pω(m|τ) at optimality; without a convergence guarantee, the context-conditioned mean field estimate carries an unquantified bias that would break the identifiability argument. A synthetic benchmark with known m could measure this bias directly.
  • Because the context space is discrete and the prior is sampled through the generative process τE ~ p, m ~ qψ(m|τE), the method's guarantees do not automatically extend to continuous context spaces; extending them would require new estimators for Eq. (8) and the partition-function sampling.
  • The taxi experiment's finding that the lost passengers are predominantly short-trip riders suggests the inferred binary context encodes trip-length preference; this could be tested by comparing inferred contexts with driver segments or with origin-destination statistics in richer data.
  • If the reward and inference model are updated alternately rather than jointly to convergence, the two desiderata (reward matching and posterior matching) may drift; treating qψ as an amortised variational posterior suggests adding an explicit variational bound as a diagnostic.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper proposes PEMMFIRL, a meta-inverse-reinforcement-learning method for mean field games (MFGs) with heterogeneous, unknown task contexts. A discrete latent context variable m with a prior p(m) is introduced; the goal is to learn a reward fω(s,a,µ,m) and an inference model qψ(m|τ) from unlabeled demonstrations generated by multiple context-conditioned MFGs, so that for a new trajectory the inferred context yields the same MFG solution as the ground-truth reward. The optimization objective combines a KL-matching term between expert and learned conditional trajectory distributions with a mutual-information maximization term, leading to a surrogate objective K(ω,ψ)−L(ω,ψ). Gradient estimators are derived by replacing the unknown mean field with an estimate constructed from qψ and by using an adversarial sampler πθ. Experiments on three simulated MFG environments and a New York City taxi-ride pricing problem are reported.

Significance. If the theoretical claims held, the paper would address a genuine gap: prior MFG IRL methods assume homogeneous agents, while meta-IRL methods handle few agents. The idea of attaching a probabilistic context variable to a family of MFGs rather than to individual agents is natural and could be useful. The experimental section covers both controlled simulations and a real-world dataset, which is a strength. However, the central derivation contains a concrete error in the mean-field estimator, and a key lemma is asserted without proof. The significance of the work is therefore conditional on repairing the theoretical core; in its current form the stated guarantees for context-free reward inference are not established.

major comments (3)
  1. [Probabilistic Embeddings for Meta-MFIRL, Eq. (8)] The claimed unbiasedness of Eq. (8) is false as written. Even if qψ(m|τ)=p(m|τ) exactly, E_{τ~pE}[qψ(m|τ)1{s_t=s}] = p(m) μ_E^t(s|m), not μ_E^t(s|m). The right-hand side is also not normalized over s (its sum is p(m)). This is not a harmless constant: pω,ψ in Eq. (9) and the κ(τ,m) term in Proposition 1 contain ∂fω/∂µ and 1/µ terms that do not cancel the scale. Consequently Propositions 1 and 2 estimate gradients of a misspecified objective. A normalized conditional estimator (e.g., divide by Σ_j qψ(m|τ_j)) is needed, or the uniform-prior case must be explicitly analyzed and the constant absorbed; neither is currently provided.
  2. [Lemma 1] Lemma 1 is load-bearing: it justifies replacing samples from the energy-based distribution pω,ψ(τ|m) with samples from πθ in the gradient estimates of Propositions 1 and 2. The lemma is stated without proof, and the claimed equality is not an immediate consequence of adversarial training. In entropy-regularized MFGs, the optimal policy is a softmax of Q-values, not of the one-step reward fω; matching pω,ψ(τ|m) requires additional conditions. The manuscript must supply a proof or a precise reference. As it stands, the gradient estimators rely on an unproved equivalence.
  3. [Eq. (10) and Algorithm 1] The generative process τE~pE(τ), m~qψ(m|τE) is claimed to match p(m) when ω and ψ are optimal. This requires qψ to equal the posterior pω(m|τ) and pω(τ|m) to equal pE(τ|m), but no fixed-point or convergence result is given for the coupled updates in Algorithm 1. Since qψ appears both in the mean-field estimator Eq. (8) and as the variational posterior, errors in qψ feed back into both the estimated mean field and the reward update. This is not a mere presentation issue; it is an unsupported assumption in the theoretical argument.
minor comments (4)
  1. [Experiments, Fig. 3 and Table 1] Fig. 3 reports median and variance over ten runs but does not show error bars or confidence intervals. Table 1 reports profit increases without standard errors or significance tests, so the real-world improvement may not be statistically distinguishable. An oracle baseline that has access to the true m would strengthen the claim that the context inference, rather than the added capacity, drives the improvement.
  2. [Appendix A] The first line of the derivation contains a typo: E_{pω}[log pω(m|τ) − qψ(m|τ)] should be E_{pω}[log pω(m|τ) − log qψ(m|τ)].
  3. [Throughout] The method name is inconsistently written as PEMMIRL in several places and PEMMFIRL elsewhere. There are also typos: 'optimility', 'nottaion', 'Pesudo-code', and 'MALW ARE'.
  4. [Taxi experiments] The taxi experiment compares only against the empirical MFG model of Ata et al.; it would be useful to also report a comparison with MF-AIRL or PLIRL on the same task, or at least explain why these baselines are not applicable.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation is an explicit variational/adversarial objective, not a definitional reduction to its inputs.

full rationale

The paper's central derivation chain is not circular. The meta-objective in Eq. (7) is obtained from explicit information-theoretic desiderata (KL matching plus mutual-information regularisation) via a written Lagrangian derivation; it is not defined as the quantity it later claims to predict. The principal suspect is Eq. (8), which defines the mean-field estimate in terms of the inference model qψ and is called 'unbiased when ψ is trained to optimality.' Even granting the skeptic's arithmetic, that expectation evaluates to p(m)µ_t^E(s|m) rather than µ_t^E(s|m); this is a false unbiasedness or estimator-identifiability claim, not a circular one. The quantity is not being defined as the target prediction; the target is the expert-conditional mean field, and the estimator is a distinct, misspecified functional of qψ. Similarly, Eq. (10) samples m from qψ(m|τE) and asserts this matches p(m) at optimality; this is a variational/bootstrap assumption, not a definitional equivalence. The use of qψ both as a variational posterior and inside the mean-field estimator is an EM-like coupling, not a case where the fitted parameter is renamed as a prediction. Self-citations to Chen et al. (2023) and Yu et al. (2019) are used as prior external results (MF-AIRL and meta-IRL with probabilistic context variables); they contain independent content and are not used as an unverified uniqueness import to forbid alternatives. Lemma 1 is stated without proof, but it is an adaptation of an existing adversarial-IRL equivalence, and its absence is a rigour gap rather than circularity. The experimental evaluation is against externally computed expert equilibria and baseline methods, so the empirical claims are not forced by construction. Overall, the paper's weaknesses are correctness and proof-gap concerns, not circular reasoning.

Assumptions & free parameters 2 free parameters · 4 assumptions · 1 invented entities

The central claim rests on the existence of a well-defined ERMFNE per context, the exactness of the variational posterior at optimality, the availability of a simulator for rollouts, and the inherited reward identifiability of MF-AIRL. The posterior exactness is the most fragile premise because the paper's own equations show the mean field estimate is not unbiased without it.

free parameters (2)
  • Context set cardinality |M|
    Experiments use |M|=2 in all tasks; qψ outputs a distribution over this fixed set. The cardinality is chosen by hand and is not inferred.
  • Lagrangian multipliers α and β = 1
    The objective in Eq. (7) fixes both to 1 with the comment that this 'confirms the incentive to maximise mutual information'. No derivation or sensitivity study is given.
assumptions (4)
  • domain assumption Existence and uniqueness of an entropy-regularised mean field Nash equilibrium for each context value m
    The expert policy and trajectory distribution are defined as an ERMFNE; the paper cites Cui and Koeppl (2021) but does not re-derive existence/uniqueness when the reward depends on the extra context variable m.
  • ad hoc to paper The variational posterior qψ(m|τ) equals the true posterior pω(m|τ) at optimality
    This is required for Eq. (8) to be unbiased and for Eq. (10) to sample from p(m). The paper gives no convergence guarantee and the derivation does not show it.
  • domain assumption A simulator is available to generate rollouts under the adaptive policy πθ
    Algorithm 1 line 7 generates trajectories using πθ and μ̂ψ, which requires sampling next states from the transition dynamics, even though the problem setup says P is not accessible.
  • domain assumption Adversarial IRL recovers the true reward up to the standard reward ambiguity
    The reward identifiability result is inherited from MF-AIRL (Chen et al. 2023) and not proven here; it is used to interpret fω as the ground-truth reward.
invented entities (1)
  • Probabilistic context variable m
    purpose: Latent variable that indexes the reward function of each structurally similar MFG task, allowing a single IRL model to handle mixed demonstrations.
    m is a modeling construct inferred from trajectories; its ground truth is set by the experimenter and there is no external falsifiable measurement validating inferred m values.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Meta-Inverse Reinforcement Learning for Mean Field Games via Probabilistic Context Variables." pith.science (2026). https://pith.science/paper/ZPP45KGV

@misc{pith2026250903845,
  author       = {Pith},
  title        = {Pith review of: Meta-Inverse Reinforcement Learning for Mean Field Games via Probabilistic Context Variables},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZPP45KGV}},
  note         = {Machine review of arXiv:2509.03845}
}
read the original abstract

Designing suitable reward functions for numerous interacting intelligent agents is challenging in real-world applications. Inverse reinforcement learning (IRL) in mean field games (MFGs) offers a practical framework to infer reward functions from expert demonstrations. While promising, the assumption of agent homogeneity limits the capability of existing methods to handle demonstrations with heterogeneous and unknown objectives, which are common in practice. To this end, we propose a deep latent variable MFG model and an associated IRL method. Critically, our method can infer rewards from different yet structurally similar tasks without prior knowledge about underlying contexts or modifying the MFG model itself. Our experiments, conducted on simulated scenarios and a real-world spatial taxi-ride pricing problem, demonstrate the superiority of our approach over state-of-the-art IRL methods in MFGs.

Figures

Figures reproduced from arXiv: 2509.03845 by the authors.

Figure 1
Figure 1. An overview of the mechanism of PEMMFIRL. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The graphic model of the meta-MFIRL problem. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Results for simulated tasks. The curves and bars [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Illustration of the initial setup. From left to right: [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Illustration for the underlying price multipliers and [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Passenger destination maps for Node 27 (left) and 52 (right). [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 40 canonical work pages

  1. [1]

    W.; Pfau, D.; Schaul, T.; Shillingford, B.; and De Freitas, N

    Andrychowicz, M.; Denil, M.; Gomez, S.; Hoffman, M. W.; Pfau, D.; Schaul, T.; Shillingford, B.; and De Freitas, N. 2016. Learning to learn by gradient descent by gradient descent. Advances in neural information processing systems, 29

    work page 2016

  2. [2]

    Ata, B.; Barjesteh, N.; and Kumar, S. 2019. Spatial pricing: An empirical analysis of taxi rides in New York City. In The University of Chicago Booth School of Business Chicago, IL Working paper

    work page 2019

  3. [3]

    Cardaliaguet, P.; and Hadikhanloo, S. 2017. Learning in mean field games: the fictitious play. ESAIM: Control, Optimisation and Calculus of Variations, 23(2): 569--591

    work page 2017

  4. [4]

    Carmona, R.; Delarue, F.; and Lachapelle, A. 2013. Control of McKean--Vlasov dynamics versus mean field games. Mathematics and Financial Economics, 7(2): 131--166

    work page 2013

  5. [5]

    Chen, Y.; Zhang, L.; Liu, J.; and Hu, S. 2022. Individual-level inverse reinforcement learning for mean field games. In Proceedings of the 21st International Conference on Autonomous Agents and Multi-agent Systems

    work page 2022

  6. [6]

    Chen, Y.; Zhang, L.; Liu, J.; and Witbrock, M. 2023. adversarial inverse reinforcement learning for mean field games. In Proceedings of the 22nd International Conference on Autonomous Agents and Multi-agent Systems

    work page 2023

  7. [7]

    Cui, K.; and Koeppl, H. 2021. Approximately Solving Mean Field Games via Entropy-Regularized Deep Reinforcement Learning. In International Conference on Artificial Intelligence and Statistics, 1909--1917. PMLR

    work page 2021

  8. [8]

    L.; Sutskever, I.; and Abbeel, P

    Duan, Y.; Schulman, J.; Chen, X.; Bartlett, P. L.; Sutskever, I.; and Abbeel, P. 2016. Rl ^2 : Fast reinforcement learning via slow reinforcement learning. arXiv preprint arXiv:1611.02779

    arXiv 2016

Show all 43 references
  1. [9]

    Elie, R.; P \'e rolat, J.; Lauri \`e re, M.; Geist, M.; and Pietquin, O. 2020. On the Convergence of Model Free Learning in Mean Field Games. In Thirty-Fourth AAAI Conference on Artificial Intelligence, 7143--7150

    2020

  2. [10]

    Finn, C.; Abbeel, P.; and Levine, S. 2017. Model-agnostic meta-learning for fast adaptation of deep networks. In International conference on machine learning, 1126--1135. PMLR

    2017

  3. [11]

    Fu, J.; Luo, K.; and Levine, S. 2018. Learning Robust Rewards with Adverserial Inverse Reinforcement Learning. In International Conference on Learning Representations

    2018

  4. [12]

    Fu, J.; Tacchetti, A.; Perolat, J.; and Bachrach, Y. 2021. Evaluating strategic structures in multi-agent inverse reinforcement learning. Journal of Artificial Intelligence Research, 71: 925--951

    2021

  5. [13]

    E.; and Hegde, N

    Ganapathi Subramanian, S.; Poupart, P.; Taylor, M. E.; and Hegde, N. 2020. Multi Type Mean Field Reinforcement Learning. In Proceedings of the 19th International Conference on Autonomous Agents and MultiAgent Systems, 411--419

    2020

  6. [14]

    Ghosh, A.; and Aggarwal, V. 2020. Model free reinforcement learning algorithm for stationary mean field equilibrium for multiple types of agents. arXiv preprint arXiv:2012.15377

    2020 arXiv

  7. [15]

    A.; Mohr, J.; and Souza, R

    Gomes, D. A.; Mohr, J.; and Souza, R. R. 2010. Discrete time, finite state space mean field games. Journal de Math \'e matiques Pures et Appliqu \'e es , 93(3): 308--328

    2010

  8. [16]

    Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; and Bengio, Y. 2014. Generative adversarial nets. Advances in neural information processing systems, 27

    2014

  9. [17]

    Guo, X.; Hu, A.; Xu, R.; and Zhang, J. 2019. Learning mean-field games. In Advances in Neural Information Processing Systems, 4967--4977

    2019

  10. [18]

    Huang, M.; and Ma, Y. 2016. Mean field stochastic games: Monotone costs and threshold policies. In 2016 IEEE 55th Conference on Decision and Control (CDC), 7105--7110. IEEE

    2016

  11. [19]

    Huang, M.; and Ma, Y. 2017. Mean field stochastic games with binary actions: Stationary threshold policies. In 2017 IEEE 56th Annual Conference on Decision and Control (CDC), 27--32. IEEE

    2017

  12. [20]

    P.; Caines, P

    Huang, M.; Malham \'e , R. P.; Caines, P. E.; et al. 2006. Large population stochastic dynamic games: closed-loop McKean-Vlasov systems and the Nash certainty equivalence principle. Communications in Information & Systems, 6(3): 221--252

    2006

  13. [21]

    Lasry, J.-M.; and Lions, P.-L. 2007. Mean field games. Japanese Journal of Mathematics, 2(1): 229--260

    2007

  14. [22]

    Mishra, N.; Rohaninejad, M.; Chen, X.; and Abbeel, P. 2017. Meta-learning with temporal convolutions. arXiv preprint arXiv:1707.03141, 2(7): 23

    2017 arXiv

  15. [23]

    Y.; Harada, D.; and Russell, S

    Ng, A. Y.; Harada, D.; and Russell, S. 1999. Policy invariance under reward transformations: Theory and application to reward shaping. In ICML, volume 99, 278--287

    1999

  16. [24]

    Y.; and Russell, S

    Ng, A. Y.; and Russell, S. J. 2000. Algorithms for Inverse Reinforcement Learning. In Proceedings of the Seventeenth International Conference on Machine Learning, 663--670

    2000

  17. [25]

    Rakelly, K.; Zhou, A.; Finn, C.; Levine, S.; and Quillen, D. 2019. Efficient off-policy meta-reinforcement learning via probabilistic context variables. In International conference on machine learning, 5331--5340. PMLR

    2019

  18. [26]

    D.; Bagnell, J

    Ratliff, N. D.; Bagnell, J. A.; and Zinkevich, M. A. 2006. Maximum margin planning. In Proceedings of the 23rd International Conference on Machine Learning, 729--736

    2006

  19. [27]

    Ravi, S.; and Larochelle, H. 2016. Optimization as a model for few-shot learning. In International conference on learning representations

    2016

  20. [28]

    Santoro, A.; Bartunov, S.; Botvinick, M.; Wierstra, D.; and Lillicrap, T. 2016. Meta-learning with memory-augmented neural networks. In International conference on machine learning, 1842--1850. PMLR

    2016

  21. [29]

    K.; Gu, S

    Seyed Ghasemipour, S. K.; Gu, S. S.; and Zemel, R. 2019. Smile: Scalable meta inverse reinforcement learning through context-conditional policies. Advances in Neural Information Processing Systems, 32

    2019

  22. [30]

    Subramanian, J.; and Mahajan, A. 2019. Reinforcement learning in stationary mean-field games. In Proceedings of the 18th International Conference on Autonomous Agents and Multi-agent Systems, 251--259

    2019

  23. [31]

    Sun, H.; Chen, X.; Shi, Q.; Hong, M.; Fu, X.; and Sidiropoulos, N. D. 2018. Learning to optimize: Training deep neural networks for interference management. IEEE Transactions on Signal Processing, 66(20): 5438--5453

    2018

  24. [32]

    Thrun, S.; and Pratt, L. 2012. Learning to learn. Springer Science & Business Media

    2012

  25. [33]

    X.; Kurth-Nelson, Z.; Tirumala, D.; Soyer, H.; Leibo, J

    Wang, J. X.; Kurth-Nelson, Z.; Tirumala, D.; Soyer, H.; Leibo, J. Z.; Munos, R.; Blundell, C.; Kumaran, D.; and Botvinick, M. 2016. Learning to reinforcement learn. arXiv preprint arXiv:1611.05763

    2016 arXiv

  26. [34]

    Y.; Benkard, C

    Weintraub, G. Y.; Benkard, C. L.; and Van Roy, B. 2010. Computational methods for oblivious equilibrium. Operations research, 58(4-part-2): 1247--1265

    2010

  27. [35]

    Xu, K.; Ratner, E.; Dragan, A.; Levine, S.; and Finn, C. 2019. Learning a prior over intent via meta-inverse reinforcement learning. In International conference on machine learning, 6952--6962. PMLR

    2019

  28. [36]

    Yang, J.; Ye, X.; Trivedi, R.; Xu, H.; and Zha, H. 2018 a . Learning Deep Mean Field Games for Modeling Large Population Behavior. In International Conference on Learning Representations

    2018

  29. [37]

    Yang, Y.; Luo, R.; Li, M.; Zhou, M.; Zhang, W.; and Wang, J. 2018 b . Mean Field Multi-Agent Reinforcement Learning. In 35th International Conference on Machine Learning, volume 80, 5571--5580. PMLR

    2018

  30. [38]

    You, C.; Lu, J.; Filev, D.; and Tsiotras, P. 2019. Advanced planning for autonomous vehicles using reinforcement learning and deep inverse reinforcement learning. Robotics and Autonomous Systems, 114: 1--18

    2019

  31. [39]

    Yu, L.; Song, J.; and Ermon, S. 2019. Multi-Agent Adversarial Inverse Reinforcement Learning. In International Conference on Machine Learning, 7194--7201

    2019

  32. [40]

    Yu, L.; Yu, T.; Finn, C.; and Ermon, S. 2019. Meta-inverse reinforcement learning with probabilistic context variables. Advances in Neural Information Processing Systems, 32

    2019

  33. [41]

    Zhao, S.; Song, J.; and Ermon, S. 2018. The information autoencoding family: A lagrangian perspective on latent variable generative models. arXiv preprint arXiv:1806.06514

    2018 arXiv

  34. [42]

    D.; Bagnell, J

    Ziebart, B. D.; Bagnell, J. A.; and Dey, A. K. 2010. Modeling interaction via the principle of maximum causal entropy. In Proceedings of the 27th International Conference on Machine Learning, 1255--1262

    2010

  35. [43]

    D.; Maas, A.; Bagnell, J

    Ziebart, B. D.; Maas, A.; Bagnell, J. A.; and Dey, A. K. 2008. Maximum entropy inverse reinforcement learning. In Proceedings of the 23rd AAAI Conference on Artificial Intelligence, 1433--1438

    2008

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.