snaykey's picture
logbook.json (verbatim titles)
200e32e verified
Raw
History Blame Contribute Delete
2.83 kB
{
"schema_version": 1,
"title": "Reproduction: Online Social Welfare Function-based Resource Allocation",
"emoji": "⚖️",
"space_id": "snaykey/repro-online-welfare-alloc",
"paper": {
"openreview_id": "qSNU4NmDpE"
},
"tags": [
"icml2026-repro",
"paper-qSNU4NmDpE"
],
"updated_at": "2026-08-02T00:00:00+00:00",
"root": {
"slug": "index",
"title": "Reproduction: Online Social Welfare Function-based Resource Allocation",
"file": "pages/index.md",
"children": [
{
"slug": "claim-1-thm41-cs-lifting",
"title": "Theorem 4.1 shows that monotonicity of the social welfare function alone suffices to lift coordinate-wise anytime-valid confidence sequences on individual utilities into a valid confidence sequence for the optimal social welfare M(mu ⊙ p*) (Theorem 4.1).",
"file": "pages/claim-1-thm41-cs-lifting/page.md",
"children": []
},
{
"slug": "claim-2-thm52-swf-ucb-regret",
"title": "Theorem 5.2 establishes that SWF-UCB attains regret R(T) = O(L(n + sqrt(nkT))) with probability 1-delta (up to polylog log log T and log(2n/delta) factors), matching lower bounds up to polylogarithmic terms (Theorem 5.2).",
"file": "pages/claim-2-thm52-swf-ucb-regret/page.md",
"children": []
},
{
"slug": "claim-3-thm51-oracles",
"title": "Theorem 5.1 provides efficient policy oracles per SWF family: an O(n log n) water-filling solution for the Weighted Power Mean and Kolm families, and an O(kn) greedy block algorithm for the Gini family (Theorem 5.1).",
"file": "pages/claim-3-thm51-oracles/page.md",
"children": []
},
{
"slug": "claim-4-sqrtt-nonmonotone-k",
"title": "Experiments across the Weighted Power Mean, Kolm, and Gini social welfare families show consistent sqrt(T) regret scaling, but regret depends non-monotonically on the resource budget k, peaking at intermediate k values (Section 6).",
"file": "pages/claim-4-sqrtt-nonmonotone-k/page.md",
"children": []
},
{
"slug": "claim-5-dependent-rounding",
"title": "Allocation feasibility under the cardinality constraint |S_t|=k is enforced via dependent rounding that preserves the marginal selection probabilities P(i in S_t) = p_{t,i} (Section 5).",
"file": "pages/claim-5-dependent-rounding/page.md",
"children": []
},
{
"slug": "executive-summary",
"title": "executive-summary",
"file": "pages/executive-summary/page.md",
"children": []
},
{
"slug": "conclusion",
"title": "conclusion",
"file": "pages/conclusion/page.md",
"children": []
}
]
}
}