1 · The essentials
These manuscripts study limits on finding nearly best answers to difficult computing problems. The main manuscript reports a resolution of the Unique Games Conjecture and has a Lean formalization. It describes an efficient way to turn a 3SAT problem into a Unique Games problem with a large gap between the two possible kinds of outcome. Four companion manuscripts address how closely algorithms can approximate answers to particular graph problems. They report barriers at the Goemans–Williamson ratio for Max-Cut, at a factor of two for Vertex Cover, and at every fixed constant factor for Min-UnCut and directed feedback vertex set. These graph results come from direct arguments rather than relying on the main Unique Games result. The Max-Cut and Vertex Cover manuscripts are also marked as Lean-formalized, while the other two are not.
A picture to help (an imperfect image, not to be taken literally)
As an imperfect image, imagine translating a puzzle into a scoring game so that a solvable puzzle permits an almost perfect score, while an unsolvable one permits only a very low score. A method that reliably distinguished those scores would also distinguish the original puzzles; the main manuscript describes a mathematical translation with this kind of gap.
Why it matters
Approximation hardness asks not just whether an exact answer is difficult to obtain, but how much accuracy remains difficult. This family reports specific barriers for several graph problems, even when the graphs have no weights. The independent direct reductions give the graph claims a route separate from the Unique Games result.
2 · Going further
The main manuscript states that, for every fixed , there is a deterministic polynomial-time reduction from 3SAT to Unique Games over a fixed finite alphabet. Its completeness is at least , and its soundness is at most . In the language of the reduction, satisfiable inputs lead to instances where almost all constraints can be satisfied, while unsatisfiable inputs lead to instances where at most a small fraction can be satisfied. The parameters are fixed in advance; the statement does not claim one fixed alphabet works for every choice of them.
The source identifies this as a resolution of Khot's Unique Games Conjecture. The central mechanism described in the abstract is the gap-producing reduction itself: it transforms the original decision problem into a sharp contrast between high and low achievable satisfaction. No construction details are supplied in the abstract, so the source does not support a more detailed account of how that gap is obtained.
The companion manuscripts report independent direct reductions using established PCP and Label Cover hardness results. For Max-Cut, the stated barrier is approximation within any fixed factor greater than the Goemans–Williamson constant, on simple unweighted graphs. For minimum Vertex Cover, it is every fixed factor below two, again on simple unweighted graphs. Both manuscripts are marked as Lean-formalized.
The remaining manuscripts claim NP-hardness of approximating Min-UnCut within every fixed factor , even on simple undirected unweighted graphs, and of approximating minimum directed feedback vertex set within any fixed constant factor, even on unweighted digraphs. These two manuscripts are not marked as Lean-formalized. Their role in the family is to extend the reported approximation barriers through direct arguments, not to supply steps needed for the main Unique Games result.
Useful notions
- Decision problems and optimization problems
- Polynomial-time reductions
- Approximation factors
- Constraint satisfaction
- Graphs and directed graphs
What was known before / what is new
Before
The source identifies the main statement as Khot's Unique Games Conjecture. It also identifies established PCP and Label Cover hardness results as inputs to the companion manuscripts' direct reductions. The supplied text gives no further historical account or comparison with earlier bounds.
New
The main manuscript reports a deterministic polynomial-time reduction from 3SAT to Unique Games with completeness arbitrarily close to one and soundness arbitrarily close to zero, for fixed parameters in the stated range. The companion manuscripts report direct approximation-hardness results for four graph problems on unweighted inputs. The main, Max-Cut, and Vertex Cover manuscripts are marked as Lean-formalized.
Limits and assumptions
The Unique Games statement requires fixed parameters and a fixed finite alphabet. The graph statements concern fixed approximation factors, with the directions of the inequalities differing between Max-Cut and the minimization problems; the stated Max-Cut and Vertex Cover claims do not include equality at their thresholds. Min-UnCut explicitly requires . The main result has a Lean formalization, and the Max-Cut and Vertex Cover manuscripts are also labelled Lean-formalized, but the supplied source does not describe the formalizations' internal scope or coverage of auxiliary claims. Min-UnCut and directed feedback vertex set are manuscript claims without a stated Lean formalization. The source lists no remaining open questions.
Mini glossary
- Polynomial-time reduction —
- A transformation from one problem to another whose running time is bounded by a polynomial in the input size.
- Completeness —
- Here, a guarantee on how large a fraction of constraints can be satisfied in the output when the original 3SAT input is satisfiable.
- Soundness —
- Here, a bound on how large a fraction of constraints can be satisfied in the output when the original 3SAT input is unsatisfiable.
- Approximation factor —
- A multiplicative measure of how close an algorithm's answer is to the best possible answer, with the convention depending on whether the objective is maximized or minimized.
- NP-hardness —
- A complexity classification expressing that a task is at least as hard as the problems in NP under the relevant reductions.
- Lean formalization —
- An encoding of mathematical statements and their proofs for checking in the Lean proof assistant.
Keywords : Unique Games · Approximation hardness · Lean formalization
AI-generated explanation based on the manuscript. Unofficial. Refer to the manuscript for the authoritative version.
3 · Technical details and sources
Claim made in the manuscript (source text)
Proves Khot's Unique Games Conjecture. Independent direct reductions also establish NP-hardness, on unweighted graphs, of approximation beyond the Goemans–Williamson ratio for Max-Cut, below factor two for Vertex Cover, and within any fixed constant factor for Min-UnCut and directed feedback vertex set. These direct proofs use established PCP and Label Cover hardness results.
Verified by Lean: the Lean software checked the part of the reasoning described below, sometimes only part of the manuscript. This does not mean independent mathematicians have validated the result.
Manuscripts in this family
The Unique Games Theorem
23/09/2026We prove the Unique Games Conjecture. For every fixed , we give a deterministic polynomial-time reduction from 3SAT to Unique Games over a fixed finite alphabet, with completeness at least and soundness at most δ.
A Direct Proof of Optimal Max-Cut Hardness
23/09/2026We prove that approximating Max-Cut on simple unweighted graphs within any fixed factor greater than the Goemans–Williamson constant is NP-hard.
The Factor-Two Hardness Threshold for Vertex Cover
23/09/2026We prove that minimum Vertex Cover is NP-hard to approximate within every fixed factor below two, even on simple unweighted graphs.
Constant-factor hardness of Min-UnCut
23/09/2026For every fixed C > 1, approximating Min-UnCut within factor C is NP-hard, even on simple undirected unweighted graphs.
Constant-factor hardness of directed feedback vertex set
23/09/2026Approximating minimum directed feedback vertex set within any fixed constant factor is NP-hard, even on unweighted digraphs.