Simultaneously Satisfying MXS and EFL

Arash Ashuri; Vasilis Gkatzelis

REVIEW 1 major objections 3 minor 1 cited by

Simultaneously Satisfying MXS and EFL

T0 review · 1 major / 3 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper proves that for a broad class of valuations, an allocation can always be simultaneously MXS and EFL, and gives an algorithm that finds one.

desk verdict A genuinely new MXS+EFL existence result with a serious but fixable flaw in the Section 8 scope proof for multiplicative valuations. read the letter →

arxiv 2412.00358 v3 pith:VFACDO5O submitted 2024-11-30 cs.GT

Arash Ashuri , Vasilis Gkatzelis This is my paper · ORCID

classification cs.GT

keywords fairdivisionindivisiblegoodsMXSEFLrestrictedMMS-feasiblevaluationsmaximinshareenvy-freeuptooneless-preferredgoodresourceallocation

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper asks whether two relaxed fairness benchmarks for indivisible goods can be met at once: MXS, an absolute benchmark based on the envy-free-up-to-any-good condition, and EFL, a relative envy-freeness notion stronger than EF1. Each was known to be achievable separately, but their simultaneous achievability was open. The paper answers this by constructing an allocation that is simultaneously MXS and EFL for every agent, and it proves the construction works for all monotone restricted-MMS-feasible valuations, a class strictly broader than additive valuations. If the argument is correct, the same allocation automatically inherits several other fairness guarantees, so one allocation covers many fairness notions at once.

What carries the argument

The central mechanism is the two-phase ReBalance subroutine, which repairs fairness after each new empty bundle is inserted into a partition. Phase 1 moves goods along envy chains between highly demanded and under-demanded bundles, and its termination is proved by a chain-dominance invariant: each repetition strictly dominates the previous allocation in a partial order on chains of the generalized envy graph. Phase 2 uses the value of a residual bundle R for agent j as a potential function: each iteration weakly increases it, and whenever the value stays the same, the bundle X_p strictly shrinks, bounding the number of iterations by the number of goods. The valuation-generalizing inequality behind both phases is restricted MMS-feasibility: for every agent, every set S, every k, and every two partitions X_k and Z_k of S, max v(X_i) >= min v(Z_i); this inequality is what lets the proof control the effect of transferring a good between bundles in Lemmas 6.7 and 6.15.

What would settle it

A concrete falsifier is a computer search over small instances: enumerate all monotone valuations on, say, four goods that satisfy Definition 2.1, run Algorithm MXS+EFL for n = 3, and test both termination and the MXS/EFL condition on the output. Finding a single restricted-MMS-feasible instance where the algorithm loops or returns an allocation that fails MXS or EFL for some agent would refute Theorem 3.1; the paper's invariants identify Lemma 6.15 as the step most likely to break first.

Watch

Extended reading notes

Core claim

Theorem 3.1 is the central claim: for any monotone restricted-MMS-feasible valuations, the MXS+EFL algorithm terminates and returns a full allocation in which every agent's bundle is both MXS-feasible and EFL-feasible. The construction is incremental: it starts with one agent holding all goods and, for k = 2 through n, adds an empty bundle and runs the ReBalance subroutine to restore MXS+EFL before proceeding. The proof that ReBalance terminates is the technical core, and it uses two separate arguments: Phase 1 uses a chain-dominance order on envy-graph chains to rule out infinite loops, and Phase 2 uses the value of a distinguished bundle for one agent as a monotone potential. Along the way the paper introduces the restricted-MMS-feasible class and shows it contains additive, budget-additive, unit-demand, and multiplicative valuations.

Load-bearing premise

The load-bearing premise is the restricted-MMS-feasible inequality of Definition 2.1: for any agent, any set of goods, and any two partitions of that set into the same number of bundles, the best bundle in one partition must be worth at least as much as the worst bundle in the other; if that inequality fails for some monotone valuation, the termination proofs in Phase 1 and Phase 2 no longer have their key comparison.

Editorial extensions

If this is right

For additive, budget-additive, unit-demand, and multiplicative valuations, an MXS+EFL allocation is guaranteed to exist and the algorithm finds one.
Every allocation returned is simultaneously EF1, 1/2-EFX, 1/2-GMMS, and 2/3-PMMS (from EFL) and 4/7-MMS (from MXS) for additive valuations, so a single allocation satisfies a broad menu of fairness notions.
The previously known best simultaneous guarantee of EFL with an alpha-MMS notion had alpha = 1/2; this result raises the MMS factor to 4/7.
Because the algorithm works for restricted-MMS-feasible valuations, the simultaneous existence result extends to a strictly broader class than additive valuations, answering an open problem raised when MXS was introduced.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The restricted-MMS-feasible condition is presented as sufficient, not necessary; if the true boundary of simultaneous MXS+EFL existence is wider, the same algorithm might run on a larger domain with only minor adjustments.
The two-phase potential-and-chain-dominance structure could be reusable for other pairs of fairness notions; the paper itself notes that EEFX+EFL cannot be handled by the same observation, so a modified invariant would be needed.
The algorithm is not claimed to be polynomial-time because FairAssociation may be implemented by exhaustive search; a natural practical extension would be to find a polynomial-time implementation for additive valuations.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Christina — quick take on arXiv:2412.00358 (Ashuri/Gkatzelis). The headline: this is a real result — an algorithm that simultaneously guarantees MXS and EFL for a broad class of monotone valuations, subsuming additive, budget-additive, and unit-demand. That combination was open; prior EFL work only got 1/2-MMS alongside, and this gets 4/7-MMS. The proof is intricate: envy-cycle elimination, chain dominance, a Phase-2 potential function. I cannot machine-check every deferred lemma, but the structure is coherent and the main termination argument is plausible. The authors also correctly observe that EFL implies EF1, 1/2-EFX, 1/2-GMMS, and 2/3-PMMS.

The soft spot is Section 8. Corollary 8.1 says multiplicative valuations are good cancelable, and therefore restricted MMS-feasible. That is false even with positive item values. Take v({d})=0.01, v({a})=10, v({b})=5; then Q={d}, R=∅, S={a}, T={b} satisfies v(Q)≥v(R) and v(S)>v(T), but v(Q∪S)=0.1 < 5 = v(R∪T). So multiplicative is not good cancelable, and the proof of Corollary 8.1 collapses. This matters because the abstract and intro advertise restricted MMS-feasible as generalizing multiplicative valuations. The claim itself is likely salvageable — a geometric-mean argument gives max over any partition ≥ (v(S))^{1/k} ≥ min over any other partition — but the paper does not supply it. Also, Lemma 8.1's proof has a case-split gap: it handles all Y bundles ≥ all X bundles and all Y bundles < all X bundles, but the mixed case is missing. Neither issue invalidates the main theorem for the restricted MMS-feasible class; it is a scope-proof problem, not a core-algorithm problem. But as written, the advertised applicability to multiplicative agents is unproven.

Verdict: worth a serious referee. The main theorem is novel, the proof strategy is a real contribution, and the Section 8 issue is fixable. I would send it to review with a request to repair the multiplicative claim (or drop it) and to add a proof sketch for the mixed case in Lemma 8.1. My own confidence in the main theorem is moderate, not high; the proof is long and unformalized, but nothing I saw points to a load-bearing error. I would cite this if I were working on EFL/MXS combinations.

Referee Report

1 major / 3 minor

Summary. The paper introduces a new class of valuation functions, called restricted MMS-feasible valuations (Definition 2.1), and presents Algorithm MXS+EFL, which it claims always terminates and returns an allocation that is simultaneously MXS and EFL for all agents whenever valuations are monotone and restricted MMS-feasible (Theorem 3.1). The algorithm iteratively grows the number of bundles from 2 to n and uses a ReBalance procedure with two phases; termination is shown through invariants, a chain-dominance relation, and a potential-function argument. Section 7 derives corollaries for additive valuations, showing that MXS allocations are 4/7-MMS and EFL allocations are 1/2-GMMS, 1/2-EFX, and 2/3-PMMS. Section 8 claims that restricted MMS-feasible valuations include good-cancelable valuations and, in particular, additive, multiplicative, budget-additive, and unit-demand valuations.

Significance. If Theorem 3.1 is correct, the paper makes a substantial contribution: it provides the first simultaneous MXS+EFL guarantee and, for additive valuations, combines approximate proportionality and approximate envy-freeness into a single allocation with several derived fairness implications. The proof strategy is original and detailed, with explicitly stated invariants and no fitted parameters; the restricted-MMS-feasible condition is an explicit domain assumption rather than a restatement of the conclusion. However, the advertised scope is not fully established: the claim that multiplicative valuations fall into the new class rests on an assertion that is false as written. The central theorem for the restricted-MMS-feasible class appears internally coherent, but the manuscript as it stands overstates its applicability to multiplicative valuations. This gap is localized and likely fixable, but it is load-bearing for the paper's stated breadth.

major comments (1)

[Section 8, Corollary 8.1] The proof that multiplicative valuations are good cancelable is invalid. Under Definition 8.1, take item values v({a})=10, v({b})=5, v({d})=0.01 and bundles Q={d}, R=∅, S={a}, T={b}. Then Q∩S=R∩T=∅, v(Q)=0.01 ≥ v(R)=0, and v(S)=10 > v(T)=5, but v(Q∪S)=0.1 < 5 = v(R∪T). Thus multiplicative valuations are not good cancelable for the paper's definition, and Corollary 8.1's assertion that multiplicative valuations are restricted MMS-feasible is unproven as written. This affects the abstract and introduction, which advertise restricted MMS-feasible as generalizing multiplicative valuations. A direct geometric-mean argument may well prove that multiplicative valuations are restricted MMS-feasible, but that argument is not supplied in the manuscript.

minor comments (3)

[Section 5.1, proof of Lemma 5.6] The sentence 'By Lemma 5.5, Ck(X′, f′) = SCk(X, f)' appears to be a typo; it should read 'SCk(X′, f′) = SCk(X, f)'.
[Section 1, Related Work] The phrase 'subsequently derived in poly-time by Chan et al.' does not give a citation year or reference number in the text; please add the appropriate citation.
[Section 8] Since multiplicative valuations are defined over arbitrary nonnegative item weights, the paper should state clearly whether the intended domain assumes all item values are at least 1; the current definition allows values below 1, which is what makes the good-cancelable claim fail.

Circularity Check

2 steps flagged · score 6.0 of 10

The Phase 2 termination proof contains two circular lemma dependencies: Lemma 6.1 and Lemma 6.2 rely on each other, and Lemma 6.10 and Lemma 6.12 form a second cycle.

other [Section 6, Proof of Lemma 6.1 using Lemma 6.2; Section 6.1, proof of Lemma 6.13]
"“Proof of Lemma 6.1 using Lemma 6.2: ReBalance updates bundle R by Secondmaxj(R, X′p, X′r) (in Line 26). By Lemma 6.3, R≺j Xp\xj_p = X′p, so we get that Secondmaxj(R, X′p, X′r)≽j R.” ... “Proof of Lemma 6.13. ... By Lemma 6.1, R′≽j R, and in transition of X to X′ only bundles Xp and Xr change.”"

Lemma 6.1 is the potential-function statement proving Phase 2 progress. Its proof explicitly invokes Lemma 6.3, whose proof relies on Phase 2 invariant A (Lemma 6.2). Lemma 6.2 is then proved inductively by Lemma 6.13, whose proof explicitly invokes Lemma 6.1. Thus the termination proof assumes the very invariant that the potential-function lemma is supposed to establish. The cycle Lemma 6.1 → Lemma 6.3 → Lemma 6.2 → Lemma 6.13 → Lemma 6.1 is circular as written.
other [Section 6.1, proofs of Lemma 6.10 and Lemma 6.12]
"“By similar arguments, by setting Y′ = (X\{Xp,Xq})∩(Z\{Za,Zb}), we get Besti(Xp,Xq)∈EFXFi(X). By Lemma 6.12, Xq∉EFXFi(X). Hence, Xp∈EFXFi(X).” ... “By Lemma 6.3 and Lemma 6.10, Xp is MXS+EFL-feasible for both agents i and j in X.”"

Lemma 6.10 concludes Xp∈EFXFi(X) by using Lemma 6.12 to rule out Xq as the best bundle. Lemma 6.12, in turn, uses Lemma 6.10 to assert that Xp is MXS+EFL-feasible for agent i. Each lemma therefore invokes the other before either is independently established. This is a second direct circular dependency in the same Phase 2 invariant-maintenance argument.

full rationale

Most of the paper is self-contained and free of the usual prediction/fitting circularity: restricted MMS-feasible valuations are an explicit domain condition, not a restatement of MXS or EFL, and the cited EFL/MXS existence results are external to the paper. The Section 8 claim that multiplicative valuations are good cancelable is not a circularity issue; it is a correctness gap, since multiplicativity fails the stated good-cancelable inequality. However, the Phase 2 termination proof as written contains two concrete circular lemma cycles. Theorem 6.1 is the central termination ingredient for Theorem 3.2 and hence for the main Theorem 3.1, so the circular dependency affects the paper's central claim as written. Score 6 reflects partial circularity in the proof, not a fitted input or self-citation chain.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No free parameters are fitted. The central claim rests on the newly introduced restricted MMS-feasible class, which is a domain assumption, plus finiteness and the oracular subroutines. No invented entities are introduced.

assumptions (3)

domain assumption Valuations are monotone and restricted MMS-feasible (Definition 2.1)
The main theorem is stated for this class; the inequality max over any partition >= min over any other partition is invoked in Lemmas 6.7, 6.15, 8.1.
standard math Finite set of goods and agents
Termination arguments count finite partitions, associations, and bundles.
domain assumption HasFairAssociation and FairAssociation subroutines can decide full MXS+EFL association existence (e.g., by exhaustive search)
The paper explicitly sets running time aside; the algorithm relies on these as oracles.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Simultaneously Satisfying MXS and EFL." pith.science (2026). https://pith.science/paper/VFACDO5O

@misc{pith2026241200358,
  author       = {Pith},
  title        = {Pith review of: Simultaneously Satisfying MXS and EFL},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VFACDO5O}},
  note         = {Machine review of arXiv:2412.00358}
}

read the original abstract

The two standard fairness notions in the resource allocation literature are proportionality and envy-freeness. If there are n agents competing for the available resources, then proportionality requires that each agent receives at least a 1/n fraction of their total value for the set of resources. On the other hand, envy-freeness requires that each agent weakly prefers the resources allocated to them over those allocated to any other agent. Each of these notions has its own benefits, but it is well known that neither one of the two is always achievable when the resources being allocated are indivisible. As a result, a lot of work has focused on satisfying fairness notions that relax either proportionality or envy-freeness. In this paper, we focus on MXS (a relaxation of proportionality) and EFL (a relaxation of envy-freeness). Each of these notions was previously shown to be achievable on its own [Barman et al.,2018, Caragiannis et al., 2023], and our main result is an algorithm that computes allocations that simultaneously satisfy both, combining the benefits of approximate proportionality and approximate envy-freeness. In fact, we prove this for any instance involving agents with valuation functions that are restricted MMS-feasible, which are more general than additive valuations. Also, since every EFL allocation directly satisfies other well-studied fairness notions like EF1, 1/2-EFX, 1/2-GMMS, and 2/3-PMMS, and every MXS allocation satisfies 4/7-MMS, the allocations returned by our algorithm simultaneously satisfy a wide variety of fairness notions and are, therefore, universally fair [Amanatidis et al., 2020].

Figures

Figures reproduced from arXiv: 2412.00358 by the authors.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

EF2X Exists For Four Agents
cs.GT 2024-11 conditional novelty 7.0 of 10

EF2X allocations are guaranteed to exist for any four-agent fair division instance with cancelable valuations, and can be computed in pseudopolynomial time.

Reference graph

Works this paper leans on

18 extracted references · 13 canonical work pages · cited by 1 Pith paper

[1]

Breaking the 3/4 barrier for ap proximate maximin share

Hannaneh Akrami and Jugal Garg. Breaking the 3/4 barrier for ap proximate maximin share. In Proceedings of the 2024 ACM-SIAM Symposium on Discrete Algorithms, SODA 2024, pages 74–91. SIAM,

work page 2024
[5]

EF2X exists fo r four agents

Arash Ashuri, Vasilis Gkatzelis, and Alkmini Sgouritsa. EF2X exists fo r four agents. In AAAI-25, Spon- sored by the Association for the Advancement of Artiﬁcial In telligence, February 25 - March 4, 2025, Philadelphia, PA, USA , pages 13555–13563. AAAI Press,

work page 2025
[9]

Fixed-point cycles and approximate EFX allocations

Benjamin Aram Berendsohn, Simona Boyadzhiyska, and L´ aszl´ o Kozma. Fixed-point cycles and approximate EFX allocations. In 47th International Symposium on Mathematical Foundations of Computer Science, MFCS 2022, volume 241 of LIPIcs, pages 17:1–17:13. Schloss Dagstuhl - Leibniz-Zentrum f¨ ur Info rmatik,

work page 2022
[11]

Maximin-aware allocations of indivisible goods

Hau Chan, Jing Chen, Bo Li, and Xiaowei Wu. Maximin-aware allocations of indivisible goods. In Proceed- ings of the 18th International Conference on Autonomous Age nts and MultiAgent Systems, AAMAS ’19, Montreal, QC, Canada, May 13-17, 2019 , pages 1871–1873. International Foundation for Autonomous Agents and Multiagent Systems. Bhaskar Ray Chaudhury, Jug...

work page 2019
[12]

Fair public dec ision making

Vincent Conitzer, Rupert Freeman, and Nisarg Shah. Fair public dec ision making. In Constantinos Daskalakis, Moshe Babaioﬀ, and Herv´ e Moulin, editors, Proceedings of the 2017 ACM Conference on Economics and Computation, EC , pages 629–646. ACM,

work page 2017
[13]

Fair allocation of indivisible goods: Improvements and generalizations

21 Mohammad Ghodsi, Mohammad Taghi Hajiaghayi, Masoud Seddighin, S aeed Seddighin, and Hadi Yami. Fair allocation of indivisible goods: Improvements and generalizations . In Proceedings of the 2018 ACM Conference on Economics and Computation , pages 539–556. ACM,

work page 2018
[14]

Rain- bow cycle number and EFX allocations: (almost) closing the gap

Shayan Chashm Jahan, Masoud Seddighin, Seyed Mohammad Seyed J avadi, and Mohammad Shariﬁ. Rain- bow cycle number and EFX allocations: (almost) closing the gap. In Proceedings of the Thirty-Second International Joint Conference on Artiﬁcial Intelligence , IJCAI 2023 , pages 2572–2580. ijcai.org,

work page 2023
[16]

Improved EFX approximation guarantees under ordinal- based assumptions

Evangelos Markakis and Christodoulos Santorinaios. Improved EFX approximation guarantees under ordinal- based assumptions. In Proceedings of the 2023 International Conference on Autono mous Agents and Multiagent Systems , pages 591–599. ACM,

work page 2023

Show all 18 references

[17]

EFX exists for three types of agents

Vishwa Prakash, Pratik Ghosal, Prajakta Nimbhorkar, and Nithin Va rma. EFX exists for three types of agents. CoRR, abs/2410.13580,

arXiv
[18]

Param eterized guarantees for almost envy-free allocations

Siddharth Barman, Debajyoti Kar, and Shraddha Pathak. Param eterized guarantees for almost envy-free allocations. In Proceedings of the 23rd International Conference on Autono mous Agents and Multiagent Systems, AAMAS 2024, Auckland, New Zealand, May 6-10, 2024 , pages 151–159...

2024
[19]

URL https://doi.org/10.48550/arXiv.2410.13580

doi: 10.48550/ARXIV.2410.13580. URL https://doi.org/10.48550/arXiv.2410.13580. 7 Implications of our Results Although our main results are stated with respect to the EFL and MX S fairness guarantees, each of these has direct implications regarding other well-studied fairness p...

doi arXiv
[61]

Hannaneh Akrami, Jugal Garg, Eklavya Sharma, and Setareh Taki

ACM, 2023a. Hannaneh Akrami, Jugal Garg, Eklavya Sharma, and Setareh Taki. Simpliﬁcation and improvement of MMS approximation. In Proceedings of the Thirty-Second International Joint Conf erence on Artiﬁcial Intelligence, IJCAI 2023 , pages 2485–2493. ijcai.org, 2023b. Georgio...

2023
[2010]

Envy-freeness up to any item with high Nash welfare: The virtue of donating items

Ioannis Caragiannis, Nick Gravin, and Xin Huang. Envy-freeness up to any item with high Nash welfare: The virtue of donating items. In Anna R. Karlin, Nicole Immorlica, and Rame sh Johari, editors, Proceedings of the 2019 ACM Conference on Economics and Computation, EC 2 019, ...

2019
[2018]

Lipton, Evangelos Markakis, Elchanan Mossel, and Amin Sa beri

Richard J. Lipton, Evangelos Markakis, Elchanan Mossel, and Amin Sa beri. On approximately fair alloca- tions of indivisible goods. In Proceedings 5th ACM Conference on Electronic Commerce (EC- 2004), pages 125–131. ACM,

2004
[2020]

Propm allocations of indivisible goods to multiple agents

Artem Baklanov, Pranav Garimidi, Vasilis Gkatzelis, and Daniel Schoep ﬂin. Propm allocations of indivisible goods to multiple agents. In Proceedings of the Thirtieth International Joint Conferen ce on Artiﬁcial Intelligence, IJCAI 2021 , pages 24–30. ijcai.org,

2021
[2023]

Pus hing the frontier on approximate EFX allocations

Georgios Amanatidis, Aris Filos-Ratsikas, and Alkmini Sgouritsa. Pus hing the frontier on approximate EFX allocations. CoRR, abs/2406.12413,

arXiv
[2024]

Epistemic EFX allocations exist for monotone valuations

Hannaneh Akrami and Nidhi Rathi. Epistemic EFX allocations exist for monotone valuations. CoRR, abs/2405.14463, 2024a. Hannaneh Akrami and Nidhi Rathi. Achieving maximin share and EFX/E F1 guarantees simultaneously. CoRR, abs/2409.01963, 2024b. Hannaneh Akrami, Rojin Rezvan, a...

2022 arXiv
[2025]

URL https://doi.org/10.1609/aaai.v39i13.33480

doi: 10.1609/AAAI.V39I 13.33480. URL https://doi.org/10.1609/aaai.v39i13.33480. Haris Aziz, Herv´ e Moulin, and Fedor Sandomirskiy. A polynomial-time a lgorithm for computing a Pareto optimal and almost proportional allocation. Oper. Res. Lett. , 48(5):573–578,

doi

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Breaking the 3/4 barrier for ap proximate maximin share

Hannaneh Akrami and Jugal Garg. Breaking the 3/4 barrier for ap proximate maximin share. In Proceedings of the 2024 ACM-SIAM Symposium on Discrete Algorithms, SODA 2024, pages 74–91. SIAM,

work page 2024

[2] [5]

EF2X exists fo r four agents

Arash Ashuri, Vasilis Gkatzelis, and Alkmini Sgouritsa. EF2X exists fo r four agents. In AAAI-25, Spon- sored by the Association for the Advancement of Artiﬁcial In telligence, February 25 - March 4, 2025, Philadelphia, PA, USA , pages 13555–13563. AAAI Press,

work page 2025

[3] [9]

Fixed-point cycles and approximate EFX allocations

Benjamin Aram Berendsohn, Simona Boyadzhiyska, and L´ aszl´ o Kozma. Fixed-point cycles and approximate EFX allocations. In 47th International Symposium on Mathematical Foundations of Computer Science, MFCS 2022, volume 241 of LIPIcs, pages 17:1–17:13. Schloss Dagstuhl - Leibniz-Zentrum f¨ ur Info rmatik,

work page 2022

[4] [11]

Maximin-aware allocations of indivisible goods

Hau Chan, Jing Chen, Bo Li, and Xiaowei Wu. Maximin-aware allocations of indivisible goods. In Proceed- ings of the 18th International Conference on Autonomous Age nts and MultiAgent Systems, AAMAS ’19, Montreal, QC, Canada, May 13-17, 2019 , pages 1871–1873. International Foundation for Autonomous Agents and Multiagent Systems. Bhaskar Ray Chaudhury, Jug...

work page 2019

[5] [12]

Fair public dec ision making

Vincent Conitzer, Rupert Freeman, and Nisarg Shah. Fair public dec ision making. In Constantinos Daskalakis, Moshe Babaioﬀ, and Herv´ e Moulin, editors, Proceedings of the 2017 ACM Conference on Economics and Computation, EC , pages 629–646. ACM,

work page 2017

[6] [13]

Fair allocation of indivisible goods: Improvements and generalizations

21 Mohammad Ghodsi, Mohammad Taghi Hajiaghayi, Masoud Seddighin, S aeed Seddighin, and Hadi Yami. Fair allocation of indivisible goods: Improvements and generalizations . In Proceedings of the 2018 ACM Conference on Economics and Computation , pages 539–556. ACM,

work page 2018

[7] [14]

Rain- bow cycle number and EFX allocations: (almost) closing the gap

Shayan Chashm Jahan, Masoud Seddighin, Seyed Mohammad Seyed J avadi, and Mohammad Shariﬁ. Rain- bow cycle number and EFX allocations: (almost) closing the gap. In Proceedings of the Thirty-Second International Joint Conference on Artiﬁcial Intelligence , IJCAI 2023 , pages 2572–2580. ijcai.org,

work page 2023

[8] [16]

Improved EFX approximation guarantees under ordinal- based assumptions

Evangelos Markakis and Christodoulos Santorinaios. Improved EFX approximation guarantees under ordinal- based assumptions. In Proceedings of the 2023 International Conference on Autono mous Agents and Multiagent Systems , pages 591–599. ACM,

work page 2023