arXiv subjects

Elie Abboud

Publications and source records attributed to Elie Abboud.

4 recordsLinked to original sources

arXiv2026

MATES: Learning Multi-Agent Interactions by Transforming Observations for Frozen Single-Agent Policies

Multi-agent reinforcement learning (MARL) commonly trains decentralized policies from scratch, requiring agents to acquire individual task competence and coordination simultaneously. Yet many multi-agent problems admit a compatible single-agent counterpart in which the underlying task can be learned in isolation. We introduce Multi-Agent Observation Transformation for Existing Single-Agent Policies (MATES), an input-side adaptation framework for tasks whose multi-agent observations preserve the solo-task information while exposing separately identifiable neighbor information. From multi-agent experience, MATES learns a small adapter that maps this observation into the format expected by a frozen single-agent policy, inducing actions suited to the shared environment without updating the single-agent policy itself. MATES leaves the pretrained policy's internal architecture unchanged and retains the objectives and update procedures of the underlying MARL algorithm. We evaluate MATES using both on- and off-policy algorithms on lifelong pathfinding, navigation, and cooperative discovery, spanning discrete and continuous observation and action spaces. Across all evaluated settings, MATES optimizes only 3.5-7.3% as many parameters as full-policy training while consistently outperforming MARL training from scratch. It approaches the performance of full fine-tuning, remains competitive overall with demonstration-based baselines, and retains strong task performance at team sizes not encountered during training. These results provide evidence that, under this observation structure, effective multi-agent behavior can be learned without modifying the policy that encodes individual competence.

cs.MA↗

arXiv2026

ARMS: Automatic Reward Shaping for Sparse-Reward Multi-Agent Reinforcement Learning

Sparse rewards are a major bottleneck in multi-agent reinforcement learning (MARL), where simultaneous learning induces non-stationarity and makes reward design especially delicate. Reward shaping can accelerate learning, but in the multi-agent setting it must preserve the strategic structure of the problem rather than merely improve short-term optimization. We propose Automatic Reward-shaping in Multi-agent Systems (ARMS), a self-supervised reward shaping framework for MARL that learns dense shaping signals from sparse environmental rewards through trajectory ranking. Since single-agent trajectory-ranking guarantees do not directly transfer to MARL, we reformulate policy invariance through conditional best-response reasoning, and show that if certain conditions hold, then using shaping rewards preserves each agent's best-response set under fixed opponent policies, and consequently preserve the set of Nash equilibria. Guided by this perspective, ARMS alternates between policy learning and reward learning while sharing shaping parameters across agents for efficiency. Experiments in a partially observable multi-agent pathfinding domain show that ARMS improves sampling efficiency under increasing reward sparsity and agent count, generalizes to unseen environments, and reveals a MARL-specific failure mode in which limited exploration and coupled policy--reward dynamics induce oscillatory behavior. Increasing exploration mitigates this effect and stabilizes learning. To the best of our knowledge, ARMS is the first automatic reward shaping framework for MARL whose design is motivated by a game-theoretic equilibrium-preservation result.

cs.MA↗

arXiv2025

Function Basis Encoding of Numerical Features in Factorization Machines

Factorization machine (FM) variants are widely used for large scale real-time content recommendation systems, since they offer an excellent balance between model accuracy and low computational costs for training and inference. These systems are trained on tabular data with both numerical and categorical columns. Incorporating numerical columns poses a challenge, and they are typically incorporated using a scalar transformation or binning, which can be either learned or chosen a-priori. In this work, we provide a systematic and theoretically-justified way to incorporate numerical features into FM variants by encoding them into a vector of function values for a set of functions of one's choice. We view factorization machines as approximators of segmentized functions, namely, functions from a field's value to the real numbers, assuming the remaining fields are assigned some given constants, which we refer to as the segment. From this perspective, we show that our technique yields a model that learns segmentized functions of the numerical feature spanned by the set of functions of one's choice, namely, the spanning coefficients vary between segments. Hence, to improve model accuracy we advocate the use of functions known to have strong approximation power, and offer the B-Spline basis due to its well-known approximation power, availability in software libraries, and efficiency. Our technique preserves fast training and inference, and requires only a small modification of the computational graph of an FM model. Therefore, it is easy to incorporate into an existing system to improve its performance. Finally, we back our claims with a set of experiments, including synthetic, performance evaluation on several data-sets, and an A/B test on a real online advertising system which shows improved performance.

cs.LG↗

arXiv2023

Finer-grained Reductions in Fine-grained Hardness of Approximation

We investigate the relation between $δ$ and $ε$ required for obtaining a $(1+δ)$-approximation in time $N^{2-ε}$ for closest pair problems under various distance metrics, and for other related problems in fine-grained complexity. Specifically, our main result shows that if it is impossible to (exactly) solve the (bichromatic) inner product (IP) problem for vectors of dimension $c \log N$ in time $N^{2-ε}$, then there is no $(1+δ)$-approximation algorithm for (bichromatic) Euclidean Closest Pair running in time $N^{2-2ε}$, where $δ\approx (ε/c)^2$ (where $\approx$ hides $\polylog$ factors). This improves on the prior result due to Chen and Williams (SODA 2019) which gave a smaller polynomial dependence of $δ$ on $ε$, on the order of $δ\approx (ε/c)^6$. Our result implies in turn that no $(1+δ)$-approximation algorithm exists for Euclidean closest pair for $δ\approx ε^4$, unless an algorithmic improvement for IP is obtained. This in turn is very close to the approximation guarantee of $δ\approx ε^3$ for Euclidean closest pair, given by the best known algorithm of Almam, Chan, and Williams (FOCS 2016). By known reductions, a similar result follows for a host of other related problems in fine-grained hardness of approximation. Our reduction combines the hardness of approximation framework of Chen and Williams, together with an MA communication protocol for IP over a small alphabet, that is inspired by the MA protocol of Chen (Theory of Computing, 2020).

cs.DS↗