rl-reward

Build RL reward signals using the OpenJudge framework. Covers choosing between pointwise and pairwise reward strategies based on RL algorithm, task type, and cost; aggregating multi-dimensional pointwise scores into a scalar reward; pairwise tournament reward for GRPO on subjective tasks (net win rate across group rollouts); generating preference pairs for DPO/RLAIF; and normalizing scores for training stability. Use when building reward models, scoring rollouts for GRPO/REINFORCE, generating preference data for DPO, or doing Best-of-N selection.

RL Reward Construction with OpenJudge

Build reward signals for reinforcement learning from human feedback (RLHF) and reinforcement learning from AI feedback (RLAIF) using the openjudge library.

When to Use This Skill

  • Building scalar rewards for GRPO / REINFORCE rollout scoring
  • Generating (chosen, rejected) preference pairs for DPO / IPO
  • Best-of-N candidate selection
  • Multi-dimensional reward shaping (correctness + safety + format)
  • Replacing or bootstrapping a reward model with LLM-as-judge

Step 1 — Choose Your Reward Strategy

Use this decision tree before writing any code:

RL Algorithm + Task type?
│
├── GRPO / REINFORCE — Verifiable task (math, code, structured output)
│   └── → POINTWISE  ✅  (FunctionGrader, exact score, zero LLM cost)
│
├── GRPO / REINFORCE — Subjective task (instruction following, dialogue, summarization)
│   └── → PAIRWISE TOURNAMENT  ✅  (compare each rollout vs all others in group,
│                                    reward = net win rate within group)
│
├── DPO / IPO / SLiC — need (chosen, rejected) pairs
│   └── → PAIRWISE  ✅  (two-way comparison, return winner/loser)
│
└── Best-of-N / reranking — rank N candidates
    └── → LISTWISE  ✅  (single call ranks all N at once)
Cost constraint?
├── Low budget
│   └── FunctionGrader (free) → pointwise; or pairwise with small judge model
│
├── Medium budget
│   └── Pointwise: 2–3 LLM graders + WeightedSumAggregator
│   └── Pairwise tournament: 1 LLM judge, N*(N-1)/2 comparisons per group
│
└── High quality / no cost limit
    └── Pointwise voting (3–5 calls) or pairwise with strong judge + debiasing

Sub-documents — Read When Relevant

TopicFileRead when…
Pointwise multi-dim rewardpointwise.mdGRPO on verifiable tasks; multi-dimension scoring
Pairwise rewardpairwise.mdGRPO on subjective tasks (tournament); DPO/RLAIF preference pairs

Read the relevant sub-document before writing any code.

Install

pip install py-openjudge

Strategy Comparison

StrategyOutputReward signalTypical useCost
Pointwisescalar per responsedirect reward r(x, y)GRPO on verifiable tasks, filteringLow–Medium
Pairwise Tournamentnet win rate per responserelative reward within groupGRPO on subjective tasksMedium (N²/2 calls)
Pairwisewinner/loser pairimplicit preference y+ > y-DPO, IPO, RLAIF preference dataMedium
Listwiserank over N responsesordinal reward / rerankingBest-of-N, rerankingMedium–High

Score Normalization

All graders return scores on different scales. Always normalize before feeding into RL:

def normalize(score: float, min_score: float, max_score: float) -> float:
    """Map [min_score, max_score] → [0.0, 1.0]."""
    if max_score == min_score:
        return 0.0
    return (score - min_score) / (max_score - min_score)

# LLM graders (common/*) return 1–5 → normalize to 0–1
reward = normalize(result.score, min_score=1, max_score=5)

# FunctionGrader / text graders already return 0–1 → no normalization needed

Evaluation Strategies

Evaluation strategies control how many times a grader is called and how results are aggregated. They are independent of the grader itself.

Choose Your Strategy

Grader type?
│
├── Deterministic (FunctionGrader, StringMatch, CodeExecution, etc.)
│   └── → Direct  (zero variance, no need for aggregation)
│
├── LLM grader — Pointwise scoring
│   │
│   ├── Budget limited / speed critical
│   │   └── → Direct  (accept variance, 1× cost)
│   │
│   ├── Discrete scores (1–5 integer, pass/fail, binary)
│   │   └── → Voting  (majority vote, robust to outliers, N× cost)
│   │
│   └── Continuous / fine-grained scores (need precise ranking)
│       └── → Average  (mean, preserves signal, N× cost)
│
└── LLM grader — Pairwise GRPO tournament
    └── → GRPOTournament  (all-pairs comparison, net win rate)
StrategyAggregationBest forCost
DirectEvaluationStrategyNoneDeterministic graders; low budget
VotingEvaluationStrategyMajority voteDiscrete / integer LLM scores
AverageEvaluationStrategyMeanContinuous LLM scores
GRPOTournamentEvaluationStrategyNet win ratePairwise GRPO on subjective tasksN²/2×

All strategies are imported from openjudge.evaluation_strategy.

Pointwise — Noise Reduction with Voting / Average

For high-variance LLM judges, wrap any grader with VotingEvaluationStrategy to run N calls and take the majority vote:

from openjudge.evaluation_strategy import VotingEvaluationStrategy

grader = CorrectnessGrader(
    model=model,
    strategy=VotingEvaluationStrategy(num_votes=3, tie_breaker="closest_to_mean"),
)
# Now each call internally runs 3 LLM evaluations and returns the most common score

Use odd num_votes (3, 5) to avoid ties.

Pairwise — GRPO Tournament

For GRPO on subjective tasks, use GRPOTournamentEvaluationStrategy to run all-pairs comparison and compute net win rate per rollout:

from openjudge.evaluation_strategy import GRPOTournamentEvaluationStrategy

strategy = GRPOTournamentEvaluationStrategy(debiased=False)
results = await strategy.execute(
    pairwise_grader.aevaluate,
    query="Write a haiku about the ocean.",
    responses=["rollout_1", "rollout_2", "rollout_3", "rollout_4"],
)
rewards = [r.score for r in results]  # net win rates in [-1.0, 1.0]

Set debiased=True to run each pair in both orders and only count consistent results (doubles LLM calls but mitigates position bias).