Pith. sign in

REVIEW 6 cited by

Stop Regressing: Training Value Functions via Classification for Scalable Deep RL

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 2403.03950 v1 pith:XPM3SDEE submitted 2024-03-06 cs.LG cs.AIstat.ML

classification cs.LGcs.AIstat.ML
keywords functionsvaluecross-entropydeepcategoricalclassificationnetworksregression
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Value functions are a central component of deep reinforcement learning (RL). These functions, parameterized by neural networks, are trained using a mean squared error regression objective to match bootstrapped target values. However, scaling value-based RL methods that use regression to large networks, such as high-capacity Transformers, has proven challenging. This difficulty is in stark contrast to supervised learning: by leveraging a cross-entropy classification loss, supervised methods have scaled reliably to massive networks. Observing this discrepancy, in this paper, we investigate whether the scalability of deep RL can also be improved simply by using classification in place of regression for training value functions. We demonstrate that value functions trained with categorical cross-entropy significantly improves performance and scalability in a variety of domains. These include: single-task RL on Atari 2600 games with SoftMoEs, multi-task RL on Atari with large-scale ResNets, robotic manipulation with Q-transformers, playing Chess without search, and a language-agent Wordle task with high-capacity Transformers, achieving state-of-the-art results on these domains. Through careful analysis, we show that the benefits of categorical cross-entropy primarily stem from its ability to mitigate issues inherent to value-based RL, such as noisy targets and non-stationarity. Overall, we argue that a simple shift to training value functions with categorical cross-entropy can yield substantial improvements in the scalability of deep RL at little-to-no cost.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

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

  1. WARP-RM: A Warp-Augmented Relative Progress Reward Model for Data Curation

    cs.RO 2026-06 unverdicted novelty 7.0 of 10

    WARP trains a reward model on time-warped successful demonstrations to produce frame-level progress estimates that upweight high-advantage chunks during behavior cloning, maintaining high success rates on suboptimal d...

  2. SERNF: Sample-Efficient Real-World Dexterous Policy Fine-Tuning via Action-Chunked Critics and Normalizing Flows

    cs.RO 2026-02 unverdicted novelty 6.0 of 10

    SERNF fine-tunes dexterous manipulation policies on real hardware by pairing normalizing-flow policies with action-chunked critics and conservative off-policy RL.

  3. ReBRAC-v2: The Return of the King

    cs.LG 2026-08 conditional novelty 5.0 of 10

    A fixed-recipe offline RL method combining normalizing-flow actors, categorical critics, staged training, and test-time refinement beats recent flow-based baselines by 22.5 points averaged over ten OGBench categories.

  4. Relative Value Learning

    cs.LG 2026-07 conditional novelty 5.0 of 10

    A critic that learns antisymmetric value differences ∆(s_i,s_j)=V(s_i)−V(s_j) has a provably contracting Bellman operator and an unbiased advantage estimator, and PPO with this critic matches standard PPO on Atari.

  5. RN-D: Discretized Categorical Actors for On-Policy Reinforcement Learning

    cs.LG 2026-01 conditional novelty 5.0 of 10

    A discretized categorical actor paired with a pre-LayerNorm residual MLP (RN-D) improves PPO's final performance and sample efficiency on continuous-control benchmarks.

  6. Value Flows

    cs.LG 2025-10 reject novelty 5.0 of 10

    Value Flows fits the full return distribution in RL with a flow-matching critic and reweights its learning objective by estimated return variance; the central theoretical guarantee does not follow from the stated equations.

Pith tools