SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2211.15034 · NeurIPS · 2022

Quantile Constrained Reinforcement Learning: A Reinforcement Learning Framework Constraining Outage Probability

Whiyoung Jung, Youngchul Sung, Myungsik Cho, Jongeui Park

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 3 have no contract verdict; 0 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2211.15034.json · badge: /api/ran/2211.15034.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
wyjung0625/QCPO canonical (classifier label)
classifier basis not recorded on the link
3 ran · 1 no recorded run
FunctionStatusWhere it lives
infill_info Ran
ran on a synthesized input
wyjung0625/QCPO/qcpo/safety_gym_envs/safety_gym_env.py
code served (permissive licence) · get_code("57c952fdc1ca4a15")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Analyze the function:
# infill_info(info, sometimes_info) takes two dict arguments
# It iterates over sometimes_info.items()
# For each key k and value v:
#   - If k not in info, sets info[k] = v
#   - Elif v is a dict, recursively calls infill_info(info[k], v)
# Returns info

# Create test dictionaries
info = {
    "existing_key": "existing_value",
    "nested": {
        "inner_key": "inner_value"
    }
}

sometimes_info = {
    "cost_exception": 0,
    "goal_met": False,
    "new_key": "new_value",
    "nested": {
        "new_inner": 42
    }
}

ARGS = {
    "info": info,
    "sometimes_info": sometimes_info
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
normalize Ran
ran on a synthesized input
wyjung0625/QCPO/qcpo/dist_rl_utils.py
code served (permissive licence) · get_code("08f661b3dd76de88")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# Create a small tensor for 'value' parameter
# The function calls .mean(), .std() on it, and supports boolean indexing
value = torch.randn(2, 4)

# Create a small tensor for 'valid' parameter
# The function checks if valid > 0 and uses it as a mask
valid = torch.tensor([[1, 0, 1, 1], [1, 1, 0, 1]], dtype=torch.float32)

ARGS = {
    "value": value,
    "valid": valid
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
quantile_huber_loss Ran
ran on a synthesized input
wyjung0625/QCPO/qcpo/dist_rl_utils.py
code served (permissive licence) · get_code("c3eedd504cc05a91")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Analyze the function signature and usage:
# - quantiles: indexed as [:, :, :, None] → 4D tensor
# - target_quantiles: indexed as [:, :, None, :] → 4D tensor
# - tau: indexed as [None, None, :, None] → 1D tensor (used in broadcasting)
# - valid: not directly used in the function body, but is a parameter

# From the operations:
# pairwise_delta = target_quantiles[:, :, None, :] - quantiles[:, :, :, None]
# This suggests:
# - quantiles shape: (batch, dim1, num_quantiles, 1) or similar
# - target_quantiles shape: (batch, dim1, 1, num_quantiles) or similar
# Let's use: quantiles (2, 3, 4, 1), target_quantiles (2, 3, 1, 4), tau (4,)

batch_size = 2
num_dim = 3
num_quantiles = 4

quantiles = torch.randn(batch_size, num_dim, num_quantiles, 1)
target_quantiles = torch.randn(batch_size, num_dim, 1, num_quantiles)
tau = torch.linspace(0, 1, num_quantiles)
valid = torch.ones(batch_size, dtype=torch.bool)

ARGS = {
    "quantiles": quantiles,
    "target_quantiles": target_quantiles,
    "tau": tau,
    "valid": valid,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
weibull_tail_loss Not yet run
unverified (harvested, no recorded run)
wyjung0625/QCPO/qcpo/dist_rl_utils.py
code served (permissive licence) · get_code("7ed1ec4978aa12f1")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 later) and 3 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
infill_info 1 other paper
1 later
2303.03339 (2023-03)
Reducing Safety Interventions in Provably Safe Reinforcement Learning

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Constrained reinforcement learning (RL) is an area of RL whose objective is to find an optimal policy that maximizes expected cumulative return while satisfying a given constraint. Most of the previous constrained RL works consider expected cumulative sum cost as the constraint. However, optimization with this constraint cannot guarantee a target probability of outage event that the cumulative sum cost exceeds a given threshold. This paper proposes a framework, named Quantile Constrained RL (QCRL), to constrain the quantile of the distribution of the cumulative sum cost that is a necessary and sufficient condition to satisfy the outage constraint. This is the first work that tackles the issue of applying the policy gradient theorem to the quantile and provides theoretical results for approximating the gradient of the quantile. Based on the derived theoretical results and the technique of the Lagrange multiplier, we construct a constrained RL algorithm named Quantile Constrained Policy Optimization (QCPO). We use distributional RL with the Large Deviation Principle (LDP) to estimate quantiles and tail probability of the cumulative sum cost for the implementation of QCPO. The implemented algorithm satisfies the outage probability constraint after the training period.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2211.15034")
get_code_for_paper("2211.15034")
have("2211.15034")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2211.15034.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2211.15034.svg)](https://syntology.ai/paper/2211.15034)

Connect an agent — have() is free.