Papers › Sequential Monte Carlo for Policy Optimization in Continuous POMDPs

Sequential Monte Carlo for Policy Optimization in Continuous POMDPs

22 May 2025arXiv:2505.16732archive 2025-07-28

Hany Abdulsamad, Sahel Iqbal, Simo Särkkä

Optimal decision-making under partial observability requires agents to balance reducing uncertainty (exploration) against pursuing immediate objectives (exploitation). In this paper, we introduce a novel policy optimization framework for continuous partially observable Markov decision processes (POMDPs) that explicitly addresses this challenge. Our method casts policy learning as probabilistic inference in a non-Markovian Feynman--Kac model that inherently captures the value of information gathering by anticipating future observations, without requiring extrinsic exploration bonuses or handcrafted heuristics. To optimize policies under this model, we develop a nested sequential Monte Carlo (SMC) algorithm that efficiently estimates a history-dependent policy gradient under samples from the optimal trajectory distribution induced by the POMDP. We demonstrate the effectiveness of our algorithm across standard continuous POMDP benchmarks, where existing methods struggle to act under uncertainty.

PaperPDFCode Syntology ranMethod as code

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 44 code samples harvested from 1 repository linked to this paper: 15 ran (of which 8 constructed an object rather than computing a result; 9 with no instrument failure: 1 honoured, 0 violated, 8 with no contract checked; 6 where Syntology's instrument failed) · 29 unverified. Where Syntology's instrument failed: 6 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: found in paper text by Syntology: 44 samples from 1 repository, 15 ran (9 with no instrument failure; 6 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written on 2026-09-22 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

Sahel13/particle-pomdp found in paper text by SyntologySyntology: 44 samples, 15 ran (6 instrument-failed) · 29 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

44 samples harvested: 15 ran (of which 8 constructed an object rather than computing a result; 9 with no instrument failure: 1 honoured, 0 violated, 8 with no contract checked; 6 where Syntology's instrument failed) · 29 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · honoured contract
6ran · our draft was wrong
8ran
29unverified

Licence: 0 of the 44 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from Sahel13/particle-pomdp. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

Sahel13/particle-pomdp found in paper text by Syntology 44 samples, 15 ran (6 instrument-failed) · 29 unverified
AttentionPolicy ppomdp/smc/_smc.py ran MIT (permissive) · 17abe752f04b80a8 · report fixture: returned AttentionPolicy · level recorded 2026-09-22
BeliefInfo ppomdp/smc/_smc.py ran MIT (permissive) · fbcd88c48e6bee65 · report fixture: returned BeliefInfo · level recorded 2026-09-22
BeliefState ppomdp/smc/_smc.py ran MIT (permissive) · 13bad645d0db5008 · report fixture: returned BeliefState · level recorded 2026-09-22
HistoryParticles ppomdp/smc/_smc.py ran MIT (permissive) · caa7f644357e4e2d · report fixture: returned HistoryParticles · level recorded 2026-09-22
HistoryState ppomdp/smc/_smc.py ran MIT (permissive) · 0eece047422502fd · report fixture: returned HistoryState · level recorded 2026-09-22
ObservationModel in 6 papers ppomdp/smc/_smc.py ran MIT (permissive) · c3c32c6451adb09a · report fixture: returned ObservationModel · level recorded 2026-09-22
RecurrentPolicy ppomdp/smc/_smc.py ran MIT (permissive) · a11c612a5c2b7ebc · report fixture: returned RecurrentPolicy · level recorded 2026-09-22
TransitionModel in 5 papers ppomdp/smc/_smc.py ran MIT (permissive) · fbce0b9594a13a87 · report fixture: returned TransitionModel · level recorded 2026-09-22
custom_split ppomdp/smc/_smc.py ran · our draft was wrong MIT (permissive) · 086446b41c921271 · report fixture: signature rng_key:ArrayImpl/1/int, num:int/0/int · returned tuple · level recorded 2026-09-22
expected_reward ppomdp/smc/_smc.py ran · our draft was wrong MIT (permissive) · dfe3485e88c1b4ff · report fixture: signature belief_state:BeliefState/0/?, action:ArrayImpl/1/float, time_idx:int/0/int, reward_fn:callable/0/? · returned ArrayImpl · level recorded 2026-09-22
log_ess ppomdp/smc/_smc.py ran · honoured contract MIT (permissive) · 3fa696d057531d54 · report fixture: signature log_weights:ArrayImpl/1/float · returned ArrayImpl · level recorded 2026-09-22
propagate_belief ppomdp/smc/_smc.py ran · our draft was wrong MIT (permissive) · 629f1ddd4988a7a5 · report fixture: signature rng_key:ArrayImpl/1/int, model:TransitionModel/1/callable, particles:ArrayImpl/2/float, action:ArrayImpl/1/float · returned ArrayImpl · level recorded 2026-09-22
reweight_belief ppomdp/smc/_smc.py ran · our draft was wrong MIT (permissive) · aa80eee012122ae9 · report fixture: signature model:ObservationModel/1/callable, state:BeliefState/1/ArrayImpl, observation:ArrayImpl/0/float · returned BeliefState · level recorded 2026-09-22
sample_marginal_obs ppomdp/smc/_smc.py ran · our draft was wrong MIT (permissive) · 6aa8ddf03cd39b4e · report fixture: signature rng_key:ArrayImpl/1/int, obs_model:ObservationModel/1/callable, state:BeliefState/1/ArrayImpl · returned ArrayImpl · level recorded 2026-09-22
systematic_resampling ppomdp/smc/_smc.py ran · our draft was wrong MIT (permissive) · c24af8993a86c4c0 · report fixture: signature rng_key:ArrayImpl/1/int, weights:ArrayImpl/1/float, num_samples:int/0/int · returned ArrayImpl · level recorded 2026-09-22
CarryAndLogProbRecurrentPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · a1ef90cedf2fc2b1 · report
EntropyAttentionPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · 05c9f57348fb470d · report
EntropyRecurrentPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · 4722555ecf97977a · report
InitializeAttentionPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · 333dfcf608c59957 · report
InitializeRecurrentPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · 1976054432381604 · report
LogProbAttentionPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · e7d2912835a24bcb · report
LogProbObservation ppomdp/smc/_smc.py unverified MIT (permissive) · 9eaf3502029a60a3 · report
LogProbRecurrentPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · 93e66e7c729b7c72 · report
LogProbTransition ppomdp/smc/_smc.py unverified MIT (permissive) · 16a27638ef41a528 · report
PathwiseCarryRecurrentPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · 21635573d5169592 · report
PathwiseLogProbRecurrentPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · 88af258e3a077113 · report
ResetAttentionPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · c5246fcd5e052c52 · report
ResetRecurrentPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · 84b66f7e86845605 · report
RewardFn ppomdp/smc/_smc.py unverified MIT (permissive) · 48c5771347c98d33 · report
SampleAndLogProbAttentionPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · 9856dc90f6891135 · report
SampleAndLogProbRecurrentPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · 307ed68087bc8258 · report
SampleAttentionPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · d74b1d12fd256ce7 · report
SampleObservation ppomdp/smc/_smc.py unverified MIT (permissive) · 6630b28821155079 · report
SampleRecurrentPolicy ppomdp/smc/_smc.py unverified MIT (permissive) · c4b2103a327f967e · report
SampleTransition ppomdp/smc/_smc.py unverified MIT (permissive) · 46a17da8fb983ce2 · report
effective_sample_size in 7 papers ppomdp/smc/_smc.py unverified MIT (permissive) · 2ee44ce129125614 · report
log_potential ppomdp/smc/_smc.py unverified MIT (permissive) · 24925fccd7fc190c · report
resample_belief ppomdp/smc/_smc.py unverified MIT (permissive) · f97fc1acfe4854ff · report
resample_history ppomdp/smc/_smc.py unverified MIT (permissive) · 3b43f9a25a70c88e · report
smc ppomdp/smc/_smc.py unverified MIT (permissive) · a21ac1f169e5e7e1 · report
smc_init ppomdp/smc/_smc.py unverified MIT (permissive) · 3e1fb6c489494efc · report
smc_step ppomdp/smc/_smc.py unverified MIT (permissive) · 71d23ef463a90bd1 · report
weighted_covar ppomdp/smc/_smc.py unverified MIT (permissive) · 4741ff263587e82f · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for nested sequential Monte Carlo algorithm (graph method: nested sequential Monte Carlo algorithm). 44 bricks; individually on a fixture: 15 ran (of which 8 constructed an object rather than computing a result; 9 with no instrument failure: 1 honoured, 0 violated, 8 with no contract checked; 6 where Syntology's instrument failed) · 29 unverified. The assembly has not been run end-to-end.

  1. smc entry · level 0 · unverified · a21ac1f169e5e7e1
  2. AttentionPolicy helper · level 1 · ran · 17abe752f04b80a8
  3. BeliefInfo helper · level 1 · ran · fbcd88c48e6bee65
  4. BeliefState helper · level 1 · ran · 13bad645d0db5008
  5. HistoryState helper · level 1 · ran · 0eece047422502fd
  6. ObservationModel helper · level 1 · ran · c3c32c6451adb09a
  7. RecurrentPolicy helper · level 1 · ran · a11c612a5c2b7ebc
  8. RewardFn helper · level 0 · unverified · 48c5771347c98d33
  9. TransitionModel helper · level 1 · ran · fbce0b9594a13a87
  10. smc_init sampler · level 0 · unverified · 3e1fb6c489494efc
  11. smc_step sampler · level 0 · unverified · 71d23ef463a90bd1
  12. systematic_resampling helper · level 1 · ran · our draft was wrong · c24af8993a86c4c0
  13. EntropyAttentionPolicy helper · level 0 · unverified · 05c9f57348fb470d
  14. InitializeAttentionPolicy helper · level 0 · unverified · 333dfcf608c59957
  15. LogProbAttentionPolicy helper · level 0 · unverified · e7d2912835a24bcb
  16. ResetAttentionPolicy helper · level 0 · unverified · c5246fcd5e052c52
  17. SampleAndLogProbAttentionPolicy helper · level 0 · unverified · 9856dc90f6891135
  18. SampleAttentionPolicy helper · level 0 · unverified · d74b1d12fd256ce7
  19. HistoryParticles helper · level 1 · ran · caa7f644357e4e2d
  20. LogProbObservation helper · level 0 · unverified · 9eaf3502029a60a3
  21. SampleObservation helper · level 0 · unverified · 6630b28821155079
  22. CarryAndLogProbRecurrentPolicy helper · level 0 · unverified · a1ef90cedf2fc2b1
  23. EntropyRecurrentPolicy helper · level 0 · unverified · 4722555ecf97977a
  24. InitializeRecurrentPolicy helper · level 0 · unverified · 1976054432381604
  25. LogProbRecurrentPolicy helper · level 0 · unverified · 93e66e7c729b7c72
  26. PathwiseCarryRecurrentPolicy helper · level 0 · unverified · 21635573d5169592
  27. PathwiseLogProbRecurrentPolicy helper · level 0 · unverified · 88af258e3a077113
  28. ResetRecurrentPolicy helper · level 0 · unverified · 84b66f7e86845605
  29. SampleAndLogProbRecurrentPolicy helper · level 0 · unverified · 307ed68087bc8258
  30. SampleRecurrentPolicy helper · level 0 · unverified · c4b2103a327f967e
  31. LogProbTransition helper · level 0 · unverified · 16a27638ef41a528
  32. SampleTransition helper · level 0 · unverified · 46a17da8fb983ce2
  33. effective_sample_size helper · level 0 · unverified · 2ee44ce129125614
  34. reweight_belief helper · level 1 · ran · our draft was wrong · aa80eee012122ae9
  35. sample_marginal_obs helper · level 1 · ran · our draft was wrong · 6aa8ddf03cd39b4e
  36. weighted_covar helper · level 0 · unverified · 4741ff263587e82f
  37. weighted_mean helper · level 0 · unverified · a27646d734705c31
  38. custom_split helper · level 1 · ran · our draft was wrong · 086446b41c921271
  39. log_potential helper · level 0 · unverified · 24925fccd7fc190c
  40. propagate_belief helper · level 1 · ran · our draft was wrong · 629f1ddd4988a7a5
  41. resample_belief helper · level 0 · unverified · f97fc1acfe4854ff
  42. resample_history helper · level 0 · unverified · 3b43f9a25a70c88e
  43. log_ess helper · level 1 · ran · honoured contract · 3fa696d057531d54
  44. expected_reward helper · level 1 · ran · our draft was wrong · dfe3485e88c1b4ff

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2505.16732") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Decision Making

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections