{"id":"9036b63a-daab-45df-a044-4c1cc7dcd92f","arxiv_id":"2505.12088","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A mod-2 intersection invariant detects nontrivial loops of embedded 2-spheres in S^2×S^2 and connected sums that evade all light-bulb moves.","lead":"This paper introduces a new counting invariant for loops of two-spheres in four-dimensional spaces, detecting loops that cannot be deformed into the simplest standard family. It is a step toward understanding whether the four-sphere carries an exotic smooth structure.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 0.3 is not established in the supplied text: the final homotopy-invariance proof is deferred to Section 7, and the completeness of the five-move classification in Theorem 6.29 is only sketched; I must be shown invariant under x3 and saddle moves, including with cross discs.","rationale":"The reader identified the completeness of the five-move classification in Theorem 6.29 as the weakest assumption. My stress-test agrees and sharpens it: even granting Theorem 6.29, the supplied text does not prove that I is invariant under the five moves, especially x3-moves and saddle moves in the multi-eye/cross-disc setting. The paper repeatedly states that invariance is proved in Section 7, but Section 7 is not provided. The rest of the architecture—the reduction from IA to EA, the switching construction, and the ordering theorem—is coherent and appears to be a genuine new method, so this is not a refutation. It is a missing proof that is essential to Theorem 0.3. Because the omission is precisely the kind of condition the reader already flagged, I do not change the CONDITIONAL verdict. The concrete check would either locate a counterexample or confirm that the deferred argument works.","tokens_in":64740,"tokens_out":7790,"duration_ms":84759,"concrete_test":"Verify the invariant under saddle moves: in the local model of §6.7.5 (Figure 46), take a finger/Whitney system before and after the saddle, with one i-disc and one j-disc, i≠j, and compute Def. 0.7 for both. If I_j changes for any such configuration, Theorem 0.3 fails. More broadly, obtain the missing §7 and check that each of the five FW-moves preserves I_j for every multi-eye system with cross discs; if Section 7 already contains this check, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Definition 0.7 defines I([α]) only after choosing a finger-first EA representative; §§2–4 show independence of choices within such representatives. The step from a fixed representative to a well-defined homotopy invariant is delegated: the text says 'In §7 we show that all such operations give a well defined invariant, thereby completing the proof' (§0), and §6 ends by identifying the FW-moves without proving their invariance. Theorem 6.29 asserts that any two finger/Whitney systems for homotopic paths differ by isotopy, disc slides, sphere slides, birth/death, x3, and saddle moves; this completeness is the linchpin connecting the local definitions to [α]. The proof of Theorem 6.29 is an outline: ordered homotopy (Thm 6.23), then a case analysis of boundary/interior intersections, cusps and saddles. In the multi-eye version, §5 gives only comments for the restandardization moves, and no computation shows that the mod-2 upper-triangular sums I_j are unchanged by x3-moves or saddle moves when cross discs are present. Since cross discs are excluded from Def. 0.7, a saddle move that changes the pairing/order of i-discs can change the sum unless a parity argument in the missing §7 rules it out. Thus the central claim is conditional on exactly the omitted invariance proof.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a codimension-2 invariant of loops of embedded 2-spheres in #^k(S^2×S^2). For a loop in finger-first position with a finger/Whitney system, it defines I_j as the mod-2 sum of intersections between ordered finger and Whitney discs pairing the j-th red and green spheres, and I=(I_1,...,I_k). The main theorem (Theorem 0.3) asserts that I is a surjective homomorphism on π_1(Emb(⊔^k S^2, #^k S^2×S^2), R_std), vanishes on loops in the light-bulb subspace, and is natural under stabilization by an extra constant sphere. Sections 2–4 develop the single-eye case and prove independence of the auxiliary choices for a fixed finger-first representative; Section 5 sketches the multi-eye extension; Section 6 analyzes generic 2-parameter families and identifies five moves relating finger/Whitney systems. The text repeatedly defers the final homotopy-invariance proof to a Section 7 that is not present in the supplied manuscript. The paper also outlines intended applications to pseudo-isotopy and to the existence of an exotic element in Diff^+(S^4) in a sequel.","tokens_in":64998,"tokens_out":5935,"duration_ms":59353,"significance":"If Theorem 0.3 is established, the result is a significant advance in simply connected 4-dimensional embedding-space theory: it produces homotopically nontrivial loops of spheres that do not come from loops of discs in light-bulb position, with a stabilization property that is absent from earlier constructions. The invariant is a direct, parameter-free geometric intersection count with no fitted constants, and the single-eye development contains a substantial and careful body of technical lemmas. Example 0.9 provides a concrete and checkable nontrivial loop. However, as submitted, the central theorem is conditional on a missing invariance proof and on the completeness of the five-move classification, so the significance can be certified only after the deferred material is supplied.","major_comments":[{"comment":"The definition of I([α]) is made after choosing a finger-first EA representative, and Sections 2–4 establish independence of choices within such representatives. The step from representatives to a well-defined homotopy invariant is deferred: the introduction states 'In §7 we show that all such operations give a well defined invariant, thereby completing the proof,' but Section 7 is not present in the supplied manuscript. Consequently Theorem 0.3 is not established in the submitted text. This is load-bearing, and the missing proof must be supplied before the central claim can be assessed.","section":"§0, Definition 0.7; §7 (missing)"},{"comment":"The completeness of the five-move description is the linchpin connecting the local invariant to the homotopy class [α], but its proof is an outline rather than a complete argument, and no invariance of I under the x3-move or the saddle move is verified here; the introduction explicitly defers this to §7. In particular, saddle moves can change the pairing or order of the discs that contribute to the mod-2 sums, and the reader is not shown why the sums are unchanged when cross discs are present. This missing verification is load-bearing for Theorem 0.3.","section":"Theorem 6.29 and §6.7"},{"comment":"The multi-eye generalization is presented as a sequence of 'Comments on' the single-eye lemmas rather than complete proofs, e.g. 'Comments on Proposition 3.24', 'Comments on Lemma 3.33', 'Comments on Lemma 3.35', 'Comments on Lemma 3.42', and 'Comments on Lemma 3.45'. Since Definition 0.7 deliberately ignores cross terms and I_j is defined eye-by-eye, one needs an explicit proof that cross discs do not affect the mod-2 upper-triangular sums under each restandardization and each FW-move. The text asserts this but does not fully demonstrate it in the multi-eye setting.","section":"§5"},{"comment":"The proof of Proposition 4.2 uses Lemma 7.9 to add a new finger and concludes I(F,W)=I(F1,W1); Remark 4.24(ii) states that this lemma is given in Section 7. Since Section 7 is absent from the submitted manuscript, Proposition 4.2 is incomplete. As Proposition 4.2 is part of the well-definedness chain for the invariant, this dependence must be supplied before the proof is complete.","section":"Remark 4.24(ii) and Proposition 4.2"}],"minor_comments":[{"comment":"The text refers to 'Theorem 6' when it appears to mean Theorem 6.23; please correct the cross-reference.","section":"§6, opening paragraph"},{"comment":"The heading 'Propositin 4.2' contains a typo and should read 'Proposition 4.2'.","section":"§5.4 heading"},{"comment":"There are small typos: 'vector feild' should be 'vector field' and 'vise versa' should be 'vice versa'.","section":"Construction 6.14 and §6.4.2"},{"comment":"The remark references Lemma 7.9 without stating it; if that lemma is to be used, its statement and proof must appear, not merely be promised for later.","section":"Remark 4.24(ii)"},{"comment":"The supplied text breaks off mid-sentence in the saddle-move subsection; if this reflects the submitted version rather than an artifact of transmission, the saddle-move analysis is incomplete and must be restored and integrated with the invariance proof.","section":"§6.7.5"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a serious research contribution and the single-eye machinery has substantial value, but the current version is incomplete: the main theorem explicitly depends on a Section 7 that is not supplied, and key multi-eye and FW-move invariance statements are deferred or only sketched. The editors should request the full Section 7 and the missing parts of Section 6.7.5 before further consideration. If the deferred proofs are supplied and are correct, the paper would be a strong candidate for publication; the current version cannot be accepted as it stands."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Mike — quick take on Gabai–Gay–Hartman. The paper introduces a mod-2 invariant I on π1 of the embedding space of k spheres in #^k S^2×S^2, built from intersections between finger and Whitney discs, and claims it detects loops that don't come from disc loops (light-bulb loops) and is stable under adding a constant sphere. That's a new and plausible route into codimension-2 embedding problems for simply connected 4-manifolds, and Example 0.9 (the f, w loop with I=1) is concrete and checkable. The detailed work in §§2–4 on disc slides, switching, restandardization, and independence of ordering is the real core of the paper, and it's done carefully, with lots of explicit lemmas and figures. The Clifford equivalence and parity arguments look sound to me at the single-eye level.\n\nThe soft spot is not small: what's in the supplied text does not prove Theorem 0.3. Well-definedness of I on π1 is explicitly deferred to §7, which is absent. The bridge, Theorem 6.29, asserts that any two finger/Whitney systems for homotopic paths differ by five specified moves; the proof given is a credible outline but not a proof, especially for the multi-eye case and for x3 and saddle moves with cross discs present. Section 5's multi-eye generalization is mostly comments on single-eye lemmas. Since cross discs are excluded from the invariant by definition, a saddle move that reorders i-discs could change the upper-triangular sum unless a parity argument in the missing §7 rules it out. That's exactly the load-bearing gap.\n\nSo the reader's conditional verdict is right. The authors are honest about the gap—they say 'In §7 we show...'—and they're not hiding it. But as an editor, you can't certify a theorem from an outline and a promise. There is enough original, detailed mathematics here to deserve referee time, and the single-eye construction may well be correct. The referee should demand the complete §7 and a fuller proof of Theorem 6.29 before publication. I wouldn't cite the main theorem yet, but I'd watch the sequel.\n\nBottom line: send it out, but condition acceptance on the missing invariance proof. This is a serious paper with a genuinely identified gap.","headline":"A genuinely new geometric invariant for loops of 2-spheres in #^k(S^2×S^2), backed by a lot of careful single-eye machinery, but the supplied text does not contain the proof of well-definedness on homotopy classes.","tokens_in":65554,"tokens_out":2519,"would_cite":false,"duration_ms":27806,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57R35","57R52","57R50","57N37"],"pacs":[],"model":"deepseek-v4-flash","headline":"A new mod-2 invariant detects loops of embedded 2-spheres in connected sums of S2×S2 that cannot be homotoped into the light-bulb space.","keywords":["Smale","4-sphere","diffeomorphism","pseudo-isotopy","loops of embedded 2-spheres","embedding spaces","Whitney discs","codimension-2"],"falsifier":"Take the loop built in Example 0.9 from a finger disc $f$ and a Whitney disc $w$ obtained by tubing the standard Whitney disc $w'$ to a 2-sphere linking $f$; the paper computes $I=1$. The theorem would be false if this loop could be homotoped, relative to its basepoint, into the light-bulb space $LB$, because then the kernel condition would force $I=0$. Equally decisive would be finding a 2-parameter deformation of finger-first loops whose finger/Whitney systems differ by an operation not among the five moves in Theorem 6.29 and for which the mod-2 count changes.","tokens_in":64527,"feed_emoji":"🔵","tokens_out":10762,"duration_ms":100299,"temperature":0.7,"pith_summary":"This paper establishes a new homotopy obstruction for loops of embedded 2-spheres in $S^2\\times S^2$ and in connected sums $\\#^k(S^2\\times S^2)$. For each $k$ it constructs a surjective homomorphism $I:\\pi_1(\\mathrm{Emb}(\\sqcup^k S^2,\\#^k(S^2\\times S^2)),R_{\\mathrm{std}})\\to\\mathbb{Z}_2^k$ that vanishes on all loops coming from the light-bulb embedding space $LB$. The $i$-th coordinate is computed by putting a loop in finger-first form and summing, mod 2, the interior intersections between ordered finger and Whitney discs in the $i$-th eye. The main work is showing that this count is unchanged under every allowable deformation of the loops and of the disc systems. If the proof is correct, the invariant gives a first codimension-2 example in simply connected 4-manifolds of sphere loops that are not disc-born and that remain nontrivial under stabilization.","feed_headline":"Invariant finds 2-sphere loops that resist homotopy","feed_subtitle":"A mod-2 count of finger and Whitney discs spots loops that are not light-bulb moves and survive stabilization.","key_machinery":"The central object is the finger/Whitney system: at the middle time of a finger-first loop, complete sets $F$ and $W$ of finger and Whitney discs pair the intersections between the moving spheres and the fixed standard spheres. The invariant is carried by the mod-2 intersection matrix $|f_p\\cap w_q|$ for $p\\le q$ after an embedded-arc normalization. To make this independent of choices the paper proves a sequence of invariance results: disc slides and switchings convert immersed arc data to embedded arc data; Clifford tori and $H_2$-equivalence absorb the ambiguity of normalizations and slides; restandardization maps cover finger twisting, braiding, spinning, and $SO(3)$-twists; and Theorem 6.29 classifies the five ways finger/Whitney systems change under generic 2-parameter homotopies: disc slides, sphere slides, birth/death moves, $x_3$-moves, and saddle moves.","core_discovery":"Theorem 0.3 asserts the existence, for every $k$, of a surjective homomorphism $I$ from $\\pi_1(\\mathrm{Emb}(\\sqcup^k S^2,\\#^k(S^2\\times S^2)),R_{\\mathrm{std}})$ to $\\mathbb{Z}_2^k$ whose kernel contains all classes represented by loops in $LB$. A loop with image $(x_1,\\dots,x_k)$ gives, upon adjoining a constant extra sphere in a new $S^2\\times S^2$ summand, a loop with image $(x_1,\\dots,x_k,0)$. Consequently there exist homotopically nontrivial loops that cannot be realized by loops of embedded 2-discs extended through the light-bulb theorem and that stay nontrivial when the number of summands is increased. The invariant is defined by representing a loop as a finger-first isotopy, collecting the finger discs $F$ and Whitney discs $W$ at the halfway time, ordering them along the immersed intersection arc, and taking the mod-2 sum $\\sum_{p\\le q}|f_p\\cap w_q|$ eye by eye.","pith_inferences":["The five-move calculus is likely useful beyond $S^2\\times S^2$: any quantity defined from finger/Whitney systems that is invariant under the five moves would produce a homotopy invariant of relative embedding classes in other closed 4-manifolds; testing this in $Y\\#^k(S^2\\times S^2)$ for a general $Y$ is a natural next step.","Because the invariant is valued in $\\mathbb{Z}_2^k$ and is stable under adding summands, it may be compatible with stabilization maps in embedding spaces; the paper records the append-zero behavior but does not determine whether $I$ factors through a stable homotopy group.","The same mod-2 counting might apply to loops of higher-genus surfaces in 4-manifolds once analogues of finger-first position and the five moves are established; the local $FW$-germ analysis in Section 6 is developed for a general surface $R$ in a general 4-manifold $X$."],"forward_implications":["The surjectivity of $I$ yields $2^k$ distinct homotopy classes of loops of $k$ spheres based at $R_{\\mathrm{std}}$, one for each vector in $\\mathbb{Z}_2^k$.","Loops with nonzero $I$ cannot be homotoped into $LB$, so they do not arise from loops of embedded 2-discs by light-bulb moves.","Stabilization preserves nontriviality: adding a constant extra sphere in a new $S^2\\times S^2$ summand appends a zero coordinate, so a nonzero class remains nonzero as $k$ increases.","For relative classes based at $LB$, Theorem 6.29 gives a complete calculus: such a class is trivial exactly when its finger/Whitney system can be reduced to the trivial system by the five $FW$-moves, which is the natural starting point for constructing smooth 4-dimensional pseudo-isotopy invariants."],"supporting_citations":[{"why":"Supplies the embedded arc condition and the finger-first ordering procedure on which the invariant's definition depends.","marker":"[Qui86]"},{"why":"Provides the light-bulb theorem used to isotope geometrically dual spheres back to $R_{\\mathrm{std}}$ and to identify the space $LB$.","marker":"[Gab20]"},{"why":"Provides twisting, Clifford tori, and restandardization tools used throughout Sections 2 through 4.","marker":"[Gab22]"},{"why":"Gives the key example of a Whitney disc tubed to a linking sphere and the problematic disc-replacement criterion that motivates the invariance analysis.","marker":"[GGHKP]"},{"why":"Provides the earlier nontrivial loop in the embedding space of a 2-sphere in $S^4$ and the related $\\pi_1(\\mathrm{Emb}(B^2,S^2\\times B^2))$ computations that $LB$ is modeled on.","marker":"[BG19]"},{"why":"Supplies the standard finger and Whitney move formalism used to describe generic isotopies and homotopies of embedded surfaces.","marker":"[FQ90]"}],"fun_headline_variants":["Mod-2 count of finger-Whitney discs reveals unhomotopable loops","New invariant detects 2-sphere loops that defy homotopy","Loops of spheres that resist light-bulb moves now detectable","Surjective invariant captures non-light-bulb loops in S^2×S^2","Finger-Whitney count finds loops that survive stabilization"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole well-definedness argument rests on Theorem 6.29 being a complete list: any two finger/Whitney systems for homotopic finger-first loops are related by isotopy and the five listed moves, a fact whose final invariance proof is deferred to Section 7.","fun_headline_variants_meta":{"raw":{"variants":["Mod-2 count of finger-Whitney discs reveals unhomotopable loops","New invariant detects 2-sphere loops that defy homotopy","Loops of spheres that resist light-bulb moves now detectable","Surjective invariant captures non-light-bulb loops in S^2×S^2","Finger-Whitney count finds loops that survive stabilization"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000551,"raw_usage":{"total_tokens":2599,"prompt_tokens":883,"completion_tokens":1716,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":499,"completion_tokens_details":{"reasoning_tokens":1618}},"tokens_in":499,"tokens_out":1716,"duration_ms":13587,"temperature":1.0,"reasoning_tokens":1618,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T20:41:43.294731+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the loop built in Example 0.9 from a finger disc $f$ and a Whitney disc $w$ obtained by tubing the standard Whitney disc $w'$ to a 2-sphere linking $f$; the paper computes $I=1$. The theorem would be false if this loop could be homotoped, relative to its basepoint, into the light-bulb space $LB$, because then the kernel condition would force $I=0$. Equally decisive would be finding a 2-parameter deformation of finger-first loops whose finger/Whitney systems differ by an operation not among the five moves in Theorem 6.29 and for which the mod-2 count changes.","supporting_citations":[],"review_version":1}