{"id":"93dff714-e73d-473d-9a42-98e4a52dc6d0","arxiv_id":"2411.19725","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In the 2D attractive Hubbard model, the polaron-to-dimeron transition present at low spin-up filling disappears above a filling of about 20%, leaving the polaron as the stable ground state at all couplings.","lead":"This paper studies a single down-spin impurity in a 2D lattice of up-spin fermions with attractive interactions, and finds that the abrupt transition between a polaron and a bound dimer state disappears once the up-spin filling is high enough. The result is relevant to cold-atom experiments with fermionic atoms in optical lattices, where the ground state at strong coupling could remain a normal Fermi liquid instead of forming pairs.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The no-transition conclusion depends on identifying E_{N↑}(k=Q_FS) with the dimeron energy; on the lattice this identification is not checked, and the sign-free PDet polaron calculation does not validate it.","rationale":"The strongest claim is that above ρ↑≈0.2 the polaron is always lower than the dimeron. The variational scan in Fig. 4 and the single PDet point at ρ≈0.29 support this, but both depend on the dimeron energy. The variational M Ansatz is an upper bound, so the PDet dimeron is the only unbiased check. That check is made with the single-particle propagator at k=Q_FS, not with the two-particle propagator, because of the sign problem. The paper justifies this by analogy to the continuum, but on the lattice the identification of the k=Q_FS sector with the dimeronic ground state is not automatic: the Fermi surface is not spherical, the band is not Galilean invariant, and the projection may converge to a finite-momentum polaron if the overlap with the bound pair is too small. I therefore agree with the reader's weakest assumption. The proposed check, scanning k or doing exact diagonalization on a small lattice, would settle whether the PDet diamonds in Fig. 2 are actually the dimeron energies. If they are not, the no-transition conclusion for ρ≳0.2 lacks unbiased numerical support, though the variational results would still suggest it. I also note that the abstract's range 0.1-0.4 conflicts with the transition visible at ρ≈0.11 in Fig. 4(b), but this is secondary to the main concern. The reader's CONDITIONAL verdict is appropriate; no verdict change is needed.","tokens_in":12688,"tokens_out":18693,"duration_ms":173214,"concrete_test":"Use the same PDet algorithm to compute E_{N↑}(k,τ→∞) from Eq. (11) for all momenta k on and around the Fermi surface at ρ≈0.29 for U/t=-10, -15, -20, and extract the residue Z(k) from the large-τ tail of G↓(k,τ). If min_k E(k) < E_{N↑}(k=0,τ→∞), or if Z(Q_FS) is negligible, the dimeron branch has been misidentified. A complementary direct check is exact diagonalization on a small lattice (e.g., 4x4 or 6x6 with closed-shell N↑ ≈ ρ L²) comparing the ground-state energy in the total-momentum sector Q_FS with the energy obtained from G↓(k=Q_FS,τ) and from the two-particle Green's function G↑↓, which tests the lattice adaptation without relying on the sign-free property.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central no-transition claim for ρ↑≳0.2 rests on the dimeron branch. At ρ≈0.29, the PDet dimeron energy in Fig. 2 is not obtained from the two-particle Green's function G↑↓ (which has a sign problem) but from the single-particle propagator G↓ at k=Q_FS via the estimator in Eq. (11). The paper imports from continuum Refs. [38,45] the statement that the polaron-dimeron transition can be viewed as a crossing of polaron states at Q=0 and |Q|=k_F. On the lattice this requires that the exact ground state in the total-momentum sector Q_FS at large |U| is the dimeronic branch, and that the trial state c†_{Q_FS,↓}|FS> has sufficient overlap with it for the large-τ projection to converge. Neither condition is demonstrated. The lattice band is not Galilean invariant, so the optimal hole-plus-pair momentum need not be Q_FS, and at U/t=-20 the overlap of the free-impurity state with a tightly bound pair may be very small, making the extracted E_D numerically unreliable. Since the variational M(Q_M=0) Ansatz is only an upper bound, the true dimeron could lie lower and cross the polaron branch exactly in the strong-coupling region where the paper claims no transition. The sign-free polaron calculation is not the issue; the single dimeron estimator is the load-bearing step.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the two-dimensional attractive Fermi-Hubbard model with one spin-down impurity and a finite filling fraction of spin-up fermions. The authors use a Chevy-type variational Ansatz with up to one particle-hole pair and a determinant diagrammatic Monte Carlo (PDet) method expanded in the bare coupling U. They report polaron and dimeron energies, quasiparticle residues, and a phase diagram as a function of U/t and spin-up filling. The central claim is that the polaron-to-dimeron transition present in the continuum and at low lattice filling disappears above a critical spin-up filling fraction close to 0.2, so that for rho_up >~ 0.2 the polaron always has lower energy and a finite residue.","tokens_in":12984,"tokens_out":11023,"duration_ms":97783,"significance":"If the central claim holds, the result is significant: it identifies lattice filling as a control parameter that can destroy the polaron-dimeron transition, with direct implications for ultracold atoms in optical lattices and for the phase diagram of the strongly spin-polarized Hubbard model. The paper also contributes a technical advance: a sign-problem-free determinant diagrammatic Monte Carlo algorithm for the impurity problem, with a direct estimator for the quasiparticle energy that avoids self-energy fitting. The demonstrated agreement between PDet and the variational polaron energy at rho=0.29 is a concrete strength, and the calculations have no fitted parameters. The main risk is not internal inconsistency of the polaron calculation but the unvalidated identification of the dimeron energy from the single-particle propagator at k=Q_FS.","major_comments":[{"comment":"The central claim is stated inconsistently. The abstract says 'we do not observe any polaron-to-dimeron transition for a range of spin-up filling fractions rho_up between 0.1 and 0.4', while the body states that the transition disappears beyond a filling fraction of about 20%. Figure 4(b), for rho_up = 101/312 ≈ 0.11, shows a transition with the crossing region magnified in the inset, and Figure 4(c), for rho_up ≈ 0.23, shows no transition. The abstract's range is therefore contradicted by the body's own data. The abstract and Fig. 4 must be reconciled; if the intended claim is 'no transition above rho_up ≈ 0.2', the phrase 'between 0.1 and 0.4' is incorrect and should be fixed.","section":"Abstract and Fig. 4"},{"comment":"The no-transition conclusion at rho_up ≈ 0.29 rests on identifying the dimeron energy with E_{N_up}(k=Q_FS, tau=infinity) obtained from the single-particle propagator G_down, not from the two-particle propagator G_updown, which the authors state suffers from a sign problem. This identification is imported from the continuum polaron-molecule literature (Refs. [38,45]). On a lattice, the band is not Galilean invariant, so the optimal momentum of the dimeronic branch need not be Q_FS, and the trial state c^dagger_{Q_FS,down}|FS> may have only a small overlap with a tightly bound pair at U/t = -20, making the large-tau projection in Eq. (10) unreliable. Since the M(Q_M=0) variational Ansatz is only an upper bound, the true dimeron could lie lower and cross the polaron branch precisely in the strong-coupling region where the paper claims no transition. The text itself notes that at small U the PDet 'dimeron' joins the bare-dimer branch, i.e. the noninteracting single-particle energy at Q_FS, which underscores that the estimator does not directly measure a two-body bound state. I ask the authors to validate the k=Q_FS estimator against an unambiguous two-particle calculation, for example G_updown at moderate |U| where the sign problem is manageable, or exact diagonalization on a small lattice, and to report the overlap of the trial state with the dimeronic branch.","section":"Results, Eq. (11), and Fig. 2"},{"comment":"The disappearance of the transition for rho_up above about 0.2 is shown only at the variational level in Fig. 4; PDet results are presented only at rho_up ≈ 0.29. Given that the PDet dimeron estimator is the least controlled element of the paper, an unbiased check at one additional filling in the critical region, such as rho_up ≈ 0.23 or 0.11, would materially strengthen the central claim. Without such a check, the statement 'Both methods give qualitatively consistent results' should be qualified as applying to the single filling where PDet was run.","section":"Fig. 4 and Conclusion"}],"minor_comments":[{"comment":"The abstract states that the algorithm is 'sign-problem free at any filling of spin-up fermions', but the body reports empirically that all sampled configurations had the same sign at the studied parameters and provides no symmetry argument. Please qualify this as an empirical observation or provide a proof.","section":"Abstract"},{"comment":"The bound argument showing that no transition can occur beyond rho_up ≈ 0.7 is only an upper bound on the critical filling; it does not by itself locate the disappearance at rho_up ≈ 0.2. Please clarify that the 0.2 threshold comes from the variational curves in Fig. 4 and not from this bound.","section":"Results, upper-bound paragraph"},{"comment":"The PDet data points in Fig. 2 appear without statistical error bars. Please add uncertainties, especially for the dimeron branch at large |U|/t, where the projection may be slow.","section":"Fig. 2"},{"comment":"The PDet algorithm and the direct energy estimator are described only briefly and partly deferred to Ref. [73], which is 'in preparation'. Please include a fuller algorithmic description or cite a preprint with the details, so that the sign-free property and the estimator can be independently reproduced.","section":"Methods, Ref. [73]"}],"recommendation":"major_revision","confidential_remarks":"The paper is potentially publishable, but the load-bearing point is the dimeron estimator. I would not require a full proof of sign-problem freeness, but a direct two-particle calculation at moderate U or exact diagonalization at small lattice sizes is needed to establish that E_{N_up}(k=Q_FS) actually gives the dimeronic branch on the lattice. The abstract/body inconsistency over the filling range must also be fixed. If the authors provide that validation, I would support publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, the paper's main claim is new: in the 2D attractive Hubbard model with a single spin-down impurity, the polaron-to-dimeron transition that exists in the continuum appears to disappear once the spin-up filling exceeds roughly 20%. Second, the Monte Carlo evidence for that disappearance is weaker than the abstract suggests, because the dimeron energy is extracted from the single-particle propagator at the Fermi momentum through an identity imported from continuum studies and never validated on the lattice.\n\nWhat's genuinely good. The determinant diagrammatic Monte Carlo (PDet) algorithm is a real technical advance: it is sign-problem free at all fillings tested, allows diagram orders above 200, and the direct energy estimator in Eq. (11) avoids the usual fitting. The polaron branch is well supported: variational and PDet energies agree at rho=0.29 up to |U|/t=20, and the quasi-particle residue stays finite, consistent with Sorella's earlier result. The variational study across fillings is systematic, and the upper-bound argument ruling out a transition beyond rho~0.7 is a useful rigorous anchor. The paper is also honest that the two-particle Green's function has a sign problem, which is why they fell back on the single-particle estimator.\n\nThe soft spots, in proportion. (1) The dimeron energy from PDet is obtained from G_down(k=Q_FS, tau), not from the two-particle propagator. On a lattice there is no Galilean invariance; the molecular ground state need not carry momentum Q_FS, and at |U|/t=20 the overlap between the free-impurity trial state and a tightly bound dimer may be tiny, making the large-tau projection unreliable. The stress-test note is right: the sign-free polaron calculation does not validate the dimeron estimator. (2) The abstract claims no transition for rho between 0.1 and 0.4, but the body identifies the critical filling as about 0.2 and Fig. 4(b) shows a transition at rho=0.11. That is a plain inconsistency. (3) The sign-free property is reported as an empirical observation, which is fine, but the phrasing \"turns out to be\" should be softened.\n\nThese issues are addressable. The core physics claim might well be correct; the variational dimer branch also shows no crossing, and the rigorous bound places an upper limit. But the decisive numerical evidence at rho=0.29 stands on the dimeron estimator, and that is a genuine weakness.\n\nI would send this to peer review. The result is significant and the polaron side is solid. A referee should ask the authors to validate the dimeron estimator, for example by exact diagonalization on small lattices or by a two-particle Green's function calculation in a regime where the sign problem is mild. If the estimator checks out, this is a valuable paper. For our reading group, I'd bring it, with the caveat.","headline":"A potentially important lattice effect, but the dimeron branch rests on an unproven estimator; the central claim needs a check before I'd trust the zero-transition conclusion.","tokens_in":13535,"tokens_out":4799,"would_cite":true,"duration_ms":37583,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"On a square lattice, the polaron-to-dimeron transition disappears once the spin-up filling fraction exceeds about 0.2.","keywords":["Fermi polaron","dimeron","attractive Hubbard model","two-dimensional lattice","diagrammatic Monte Carlo","variational ansatz","quasi-particle residue","polaron-to-dimeron transition"],"falsifier":"Compute the dimeron energy directly from the two-particle Green's function G_updown(r,tau) on the same lattice and filling, for example rho_up around 0.29, using a method that can handle its sign-changing diagrams, and compare it with the polaron energy from G_down(k=0,tau). If the direct dimeron energy dips below the polaron energy for |U|/t greater than 6 at such filling, the paper's central claim of no transition above rho_up around 0.2 is contradicted. A simpler indirect check is to repeat the present PDet extraction at several larger L and N_up values to confirm that the PDet dimeron data points in Fig. 2 saturate; a strong system-size drift would signal that the k=Q_FS proxy is unreliable.","tokens_in":12434,"feed_emoji":"⚛️","tokens_out":5392,"duration_ms":43497,"temperature":0.7,"pith_summary":"This paper asks whether the polaron-to-dimeron transition seen in a two-dimensional continuum Fermi gas survives when the same system is placed on a square optical lattice. The authors find that it does at low spin-up filling, but the transition shifts to stronger attraction as filling rises and disappears above a critical spin-up filling fraction of about 0.2. For a single spin-down impurity in the attractive Hubbard model, the polaron, an impurity dressed by particle-hole excitations, then remains the ground state at all couplings and keeps a finite quasi-particle residue. This matters because it changes the predicted phase diagram of strongly imbalanced fermions in optical lattices: dilute spin-down impurities should form a superfluid at strong coupling only at low spin-up filling, while above the critical filling the system should stay a normal Fermi liquid.","feed_headline":"Lattice kills the polaron-dimeron transition above 20% filling","feed_subtitle":"In the 2D attractive Hubbard model, a single impurity stays a polaron at any coupling once spin-up density passes ~0.2.","key_machinery":"The argument is carried by two complementary tools. The first is the variational wave functions |P(Q_P)> and |M(Q_M)>, the polaron with up to one particle-hole pair and the dimeron with up to two particle-hole pairs, whose energy is obtained by solving a Fredholm equation for the kernel determinant. The second is a polaron determinant (PDet) diagrammatic Monte Carlo algorithm, in which the sum of all diagram topologies at given interaction-vertex coordinates is written as a single determinant of non-interacting propagators; this expansion is found to be sign-problem-free at any spin-up filling for the polaron propagator. A key identification is that the dimeron energy is extracted from the polaron propagator at momentum k=Q_FS, the Fermi momentum, following the continuum observation that the polaron-to-dimeron transition is a crossing of the Q=0 and |Q|=k_F polaron branches. The quasi-particle residue is read off the large-time decay of the polaron propagator.","core_discovery":"Working with the two-dimensional attractive Fermi-Hubbard model at zero temperature with one spin-down impurity in a Fermi sea of spin-up fermions, the paper establishes that the sharp first-order transition between a polaron and a dimeron, a dressed bound pair, is not a robust feature of the lattice problem. At low spin-up filling the transition is present and approaches the continuum result, but as the filling fraction is increased the critical |U|/t grows, and beyond a filling of roughly 0.2 the transition is absent for |U|/t up to 20 and, via the variational states, for all couplings studied. In this regime the polaron energy is always lower than the dimeron energy and the polaron residue stays finite. The result is supported by two independent methods: a variational ansatz truncated at one particle-hole excitation, and a determinant diagrammatic Monte Carlo algorithm that samples the bare-U expansion to orders beyond 200 without a sign problem; a direct estimator extracts the ground-state energy from the propagator without fitting.","pith_inferences":["The sign-problem-free property of the polaron determinant series at arbitrary spin-up filling is not explained by any symmetry in the paper; if it holds for a finite density of spin-down fermions rather than a single impurity, the algorithm could extend to the full strongly polarized Hubbard model phase diagram.","The upper-bound argument places the rigorous ceiling for the transition at rho_up around 0.7, well above the observed 0.2; narrowing this gap with improved variational states or a direct dimeron propagator would either strengthen or revise the disappearance claim.","The same determinant Monte Carlo approach could be applied to other lattice geometries or to mass-imbalanced impurities, where the bandwidth and Fermi-surface shape might restore or further suppress the transition.","If the polaron remains stable with finite residue at all couplings for rho_up above 0.2, the fate of a dilute gas of impurities is a normal Fermi liquid rather than a superfluid, suggesting a filling-controlled quantum phase transition in the strongly polarized Hubbard model that could be mapped experimentally by measuring the impurity spectral function."],"forward_implications":["At spin-up filling below about 0.2, the polaron-to-dimeron transition exists and shifts to larger |U|/t as filling increases, matching the continuum two-dimensional result in the low-filling limit.","Above rho_up around 0.2 in the attractive Hubbard model, a single spin-down impurity remains a polaron at all couplings, and the dimeronic branch never becomes the ground state.","The polaron quasi-particle residue Z_0 remains finite for all couplings at rho_up around 0.29, consistent with an earlier lattice polaron calculation.","In cold-atom experiments with two-dimensional optical lattices, a small density of spin-down impurities should form a superfluid at strong coupling only at low spin-up filling; above the critical filling the strongly polarized gas should remain a normal Fermi liquid.","The new determinant diagrammatic Monte Carlo estimator gives polaron energies directly without self-energy fitting, and the algorithm samples diagram orders above 200 with fixed sign even at U/t=-20."],"supporting_citations":[{"why":"Supplies the continuum two-dimensional polaron-dimeron transition that the paper seeks on the lattice and recovers at low filling.","marker":"[31]"},{"why":"Motivates extracting the dimeron energy from the polaron propagator at the Fermi momentum.","marker":"[38]"},{"why":"Further establishes the Q=0 versus |Q|=k_F crossing and the unified variational treatment of both branches.","marker":"[45]"},{"why":"Provides the earlier lattice Hubbard calculation whose finite polaron residue the paper confirms.","marker":"[48]"},{"why":"Supplies the determinant-diagrammatic algorithm that the paper adapts and extends with a direct energy estimator.","marker":"[69]"},{"why":"Introduces the one-particle-hole variational ansatz used to construct the polaron and dimeron wave functions.","marker":"[33]"}],"fun_headline_variants":["Polaron always wins above 20% filling in 2D Hubbard","No dimeron transition for rho>0.2 in 2D attractive Hubbard","Lattice prevents polaron-dimeron transition at high density","Polaron stable: no transition above 20% filling"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The dimeron ground-state energy is obtained from the single-particle polaron propagator evaluated at the Fermi momentum, a relation carried over from continuum studies; if this momentum-to-dimeron identification is not accurate at finite lattice filling, the reported disappearance of the transition could be an artifact of the energy estimate.","fun_headline_variants_meta":{"raw":{"variants":["Polaron always wins above 20% filling in 2D Hubbard","No dimeron transition for rho>0.2 in 2D attractive Hubbard","Lattice prevents polaron-dimeron transition at high density","Polaron stable: no transition above 20% filling"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000608,"raw_usage":{"total_tokens":2866,"prompt_tokens":1016,"completion_tokens":1850,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":632,"completion_tokens_details":{"reasoning_tokens":1771}},"tokens_in":632,"tokens_out":1850,"duration_ms":11523,"temperature":1.0,"reasoning_tokens":1771,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T05:53:55.903192+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the dimeron energy directly from the two-particle Green's function G_updown(r,tau) on the same lattice and filling, for example rho_up around 0.29, using a method that can handle its sign-changing diagrams, and compare it with the polaron energy from G_down(k=0,tau). If the direct dimeron energy dips below the polaron energy for |U|/t greater than 6 at such filling, the paper's central claim of no transition above rho_up around 0.2 is contradicted. A simpler indirect check is to repeat the present PDet extraction at several larger L and N_up values to confirm that the PDet dimeron data points in Fig. 2 saturate; a strong system-size drift would signal that the k=Q_FS proxy is unreliable.","supporting_citations":[{"cited_title":"Vlietinck, J","cited_arxiv_id":null,"evidence_quote":"Supplies the continuum two-dimensional polaron-dimeron transition that the paper seeks on the lattice and recovers at low filling."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Motivates extracting the dimeron energy from the polaron propagator at the Fermi momentum."},{"cited_title":"Cui, Fermi polaron revisited: Polaron-molecule tran- sition and coexistence, Phys","cited_arxiv_id":null,"evidence_quote":"Further establishes the Q=0 versus |Q|=k_F crossing and the unified variational treatment of both branches."},{"cited_title":"Sorella, Numerical evidence of Luttinger- and Fermi- liquid behavior in the two-dimensional Hubbard model, Phys","cited_arxiv_id":null,"evidence_quote":"Provides the earlier lattice Hubbard calculation whose finite polaron residue the paper confirms."},{"cited_title":"Van Houcke, F","cited_arxiv_id":null,"evidence_quote":"Supplies the determinant-diagrammatic algorithm that the paper adapts and extends with a direct energy estimator."},{"cited_title":"Chevy, Universal phase diagram of a strongly interact- ing Fermi gas with unbalanced spin populations, Phys","cited_arxiv_id":null,"evidence_quote":"Introduces the one-particle-hole variational ansatz used to construct the polaron and dimeron wave functions."}],"review_version":1}