Methods › Reinforcement Learning › Value Function Estimation › V-trace

V-trace

34 papers tagged archive 2025-07-28

Introduced by Lasse Espeholt et al. in IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures

archive 2025-07-28 Description, source and code snippet are the archive's method entry.

V-trace is an off-policy actor-critic reinforcement learning algorithm that helps tackle the lag between when actions are generated by the actors and when the learner estimates the gradient. Consider a trajectory (xₜ, aₜ, rₜ)ᵗ⁼ˢ⁺ⁿₜ₌ₛ generated by the actor following some policy μ. We can define the n-steps V-trace target for V(xₛ), our value approximation at state xₛ as:

vₛ = V(xₛ) + ∑ˢ⁺ⁿ⁻¹ₜ₌ₛγᵗ⁻ˢ(∏ᵗ⁻¹ᵢ₌ₛcᵢ)δₜV

Where δₜV = ρₜ(rₜ + γV(xₜ₊₁) - V(xₜ)) is a temporal difference algorithm for V, and ρₜ = min(ρ̅, (π(aₜ|xₜ))/(μ(aₜ|xₜ))) and cᵢ = min(c̅, (π(aₜ|xₜ))/(μ(aₜ|xₜ))) are truncated importance sampling weights. We assume that the truncation levels are such that ρ̅ ≥c̅.

PaperSource

In Syntology code not checked for this name. The lookup matched a different name in Syntology's graph, 'VTrace' (resolution: token_match), and its answer is about that name only, so it does not say whether this method has code. About another paper: this answer is about arXiv:2609.37656, while the archive credits arXiv:1802.01561.

Code Syntology holds under other names, as the lookup lists them (code_nearby); not a complete list, and none of them is this method.

Syntology's have check for this name (free, it answers coverage only), read 2026-10-05; ask it again live. A level is the served verification level of that one answer; nothing here is summed across methods.

What each level means, quoted from Syntology's server card: V0 “imports”; V1 “executes against a synthesized fixture and returns a value of the recorded type”; V2 “two implementations GENERATED from the paper's signature and spec agree numerically (1e-6) on fixture inputs” “neither of the two ever read the authors' own code, so V2 measures how unambiguously the paper specifies the method, not fidelity to the authors' implementation (measured 2026-09-11, commit 75eaa77: cross_checked_by is a second LLM on 2,628 of 2,831 served samples)”; V3 “paper-derived property tests pass on both of those same two generations.” The card adds: “No rung on this ladder compares the served code against the authors' implementation, so treat every level as evidence about runnability and internal agreement, not about matching the paper's results.” A level quoted in two parts is one definition the card writes as two clauses joined by a double hyphen, each clause quoted on its own (how these levels are quoted).

Papers archive 2025-07-28

archive papers tagged: 34 · with a code link: 19 · where Syntology ran a sample: 8 (7 with a run with no instrument failure, 1 where every run was a failure of Syntology's instrument) Syntology

Show: all tagged papersonly where code ran (8 of 34 tagged: 7 with a run with no instrument failure, 1 where every run was a failure of Syntology's instrument)

Page 1 of 1: papers 1 to 34 of 34, newest first by the archive's date (ties by slug), in archive order.

Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code, as “N ran (of which C constructed an object rather than computing a result; K with no instrument failure: H honoured, V violated, P with no contract checked; I where Syntology's instrument failed) · U unverified”; the instrument figure counts failures of Syntology's instrument, not of the code. It is per sample and not a correctness claim. When the archive marks a repository official for the paper, the line starts with that repository's state (the archive's flag, not a verdict on who wrote the code; “community repositories only” when every sample that ran came from a community repository, “official: no sample here; runs from other or unrecorded repositories” when some came from a repository the paper names or has in its text, or from none recorded); hover it for the repositories the samples that ran came from.

Tasks archive 2025-07-28

20 shown of 31 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.

TaskPapers
Reinforcement Learning18
Reinforcement Learning (RL)15
reinforcement-learning14
Starcraft7
Starcraft II7
Deep Reinforcement Learning5
Atari Games3
Decision Making3
Continuous Control2
Lifelong learning2
OpenAI Gym2
Q-Learning2
Transfer Learning2
continuous-control2
AI Agent1
CPU1
Diagnostic1
Diversity1
Edge-computing1
Efficient Exploration1

Usage over time archive 2025-07-28

Papers per year tagged with V-trace: 2018 to 2025, peak 9 9 0 2018: 1 paper 2018 2019: 6 papers 2019 2020: 4 papers 2020 2021: 9 papers 2021 2022: 5 papers 2022 2023: 7 papers 2023 2024: 1 paper 2024 2025: 1 paper 2025
Papers per year the archive tags with this method, by the paper's archive date (34 dated). Bars are counts, not a trend claim.

Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).

Categories archive 2025-07-28

Value Function Estimation

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