{"id":"848956cd-7fe3-4ee5-ae2c-76eb2cb07eda","arxiv_id":"2505.07356","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Active ultrastable glasses exhibit an inverse Bauschinger effect: after a small pre-shear, they yield at higher stress in the reverse direction, then revert to classical softening at larger strains or under cyclic shear.","lead":"This paper uses computer simulations to show that a special type of glass with self-propelled particles can briefly become stronger, not weaker, when the direction of squeezing is reversed. This memory-like behavior could let engineers design materials that store or erase deformation history on demand.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Missing control for activity-induced aging: the reverse and forward tests take different times, so 'healing upon shear reversal' may be a waiting-time artifact rather than directional memory.","rationale":"The central claim has two parts: (i) an observation that reverse yield stress exceeds forward yield stress after pre-shear, and (ii) a mechanism in which shear reversal transiently heals shear band networks. The observation is made on single stress-strain curves with no error bars or stated system size, so part (i) is statistically unverified, as the reader noted. But part (ii) has a more specific hidden assumption: in an actively driven system, time and strain are confounded. Because the reverse test reaches yield at larger strain, it runs longer, giving active particles more time to reorganize the shear band network. The paper's Fig. 4 shows healing during reverse shear but no zero-shear aging control, so it cannot distinguish direction-induced healing from activity-induced relaxation. This is not a disagreement with consensus; it is a missing control in the paper's own protocol. A single aging control would settle it. If the control shows no forward/reverse asymmetry after equal waiting, the claimed inverse Bauschinger effect would not be a mechanical memory effect; if the control shows forward yield still lower, the directional mechanism survives. I would keep the reader's CONDITIONAL verdict, since the paper otherwise has a coherent story and useful controls (fresh-sample symmetry, activity switch-off), but the acceptance conditions should include this control and ensemble averaging.","tokens_in":15463,"tokens_out":9661,"duration_ms":102365,"concrete_test":"From the same zero-stress configuration used in Fig. 3(c) (γN = 0.13, τp = 0.1, f0 = 2.0), perform an activity-only aging run at the same state for a duration equal to the reverse-shear path from γ = 0.024 to the reverse yield peak (≈ γ = -0.08 at γdot = 5e-5), then load in the forward direction. Repeat on at least 10 independent configurations. If the aged forward yield stress matches the original reverse value, the inverse Bauschinger effect is an aging artifact; if it stays at the original forward value, directional healing is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"At the zero-stress state used for Fig. 3(c) (γN = 0.13, τp = 0.1, f0 = 2.0, γdot = 5e-5), the active system is still being driven: the shear band network can reorganize even without shear. The forward and reverse tests are separate runs from this state, and the reverse branch yields at a larger strain, hence after a longer simulated time. If activity alone heals or weakens the band over that interval, the elevated reverse yield stress would be a waiting-time artifact, not evidence of directional mechanical memory. The paper's claim that 'the system heals upon shear reversal' conflates time under activity with reversal of strain direction. The activity-switch-off experiment in Supplementary Fig. S1 shows the active state is time-evolving, and no zero-shear aging control is reported. Statistical averaging is also absent (single curves, no system size or error bars), but the decisive missing check is this aging control.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents molecular dynamics simulations of a two-dimensional polydisperse soft-sphere ultrastable glass doped with 20% run-and-tumble active particles, subjected to shear deformation with direction reversal at various pre-shear strains. The authors report that, unlike the passive glass which shows the classical Bauschinger effect, the active glass at low persistence time exhibits an inverse Bauschinger effect for intermediate pre-shear strains: after unloading to a zero-stress state, the yield stress under reverse shear exceeds that under forward shear. The effect is attributed to the transient healing of a network of shear bands during reverse loading, and the system is reported to cross over to the classical Bauschinger effect at larger pre-shear strains. The paper also studies oscillatory shear, reporting progressive shear softening and irreversible restructuring of shear band networks in active systems.","tokens_in":15639,"tokens_out":3083,"duration_ms":32830,"significance":"The claimed observation, if statistically robust, would be the first report of an inverse Bauschinger effect in an amorphous solid, and it would connect active driving to mechanically encoded directional memory in ultrastable glasses. The study has notable strengths: the effect is extracted directly from simulated stress-strain curves with no fitted parameters, a systematic scan over active force magnitude, persistence time, and pre-shear strain is presented, and the fabric-tensor anisotropy analysis provides a plausible microstructural correlate. The shear-band healing mechanism is visually supported by D2min snapshots and movies. However, the central quantitative claim rests on single realizations without reported system-size dependence or ensemble averaging, and the protocol lacks a control for activity-induced aging during the reverse test, so the significance claim is currently conditional on closing these gaps.","major_comments":[{"comment":"The central claim of an inverse Bauschinger effect is based on the relative heights of forward and reverse yield peaks in Fig. 3, but the paper does not report the system size, the number of independent configurations, or any ensemble averaging or error bars. The stress-strain curves appear to be single realizations. Since shear band formation in small two-dimensional samples is strongly realization-dependent, the reported asymmetry between forward and reverse yield stress at γN = 0.13 could be within sample-to-sample fluctuations. Please state the system size and number of independent samples and show averaged stress-strain curves with standard errors, and provide the numerical yield stresses for the states in Fig. 3(b-f) with an explicit operational definition of yield stress.","section":"Models and Methods; Section II, Fig. 3(b-f)"},{"comment":"The inverse Bauschinger claim conflates time under activity with the reversal of strain direction. At the zero-stress state used for Fig. 3(c) (γN = 0.13, τp = 0.1, f0 = 2.0), the active forces are still applied and the system continues to evolve even without shear. The forward and reverse tests are separate runs starting from this state, and the reverse branch yields at a larger strain and therefore after a longer simulated time. If activity alone heals or weakens the shear band network over that interval, the elevated reverse yield stress would be a waiting-time artifact rather than evidence of directional mechanical memory. The authors must perform a control in which the same zero-stress state is aged under activity for different waiting times before shear is applied in either direction, and show that the yield stress difference is not reproduced by aging alone. The claim that 'the system heals upon shear reversal' requires such a control to distinguish direction-dependent healing from simple time-dependent relaxation.","section":"Section II, Fig. 3(c) and 'Healing of Shear Band Networks'"},{"comment":"The manuscript's central novelty claim, 'we believe that our observation of the inverse Bauschinger effect in amorphous solids could be the first documented observation', is stronger than the evidence currently supports. A literature statement that, to the authors' knowledge, no previous simulation or experiment on amorphous solids has reported an inverse Bauschinger effect should be supported by a dedicated comparison with prior work on metallic glasses, polymer glasses, and computational amorphous solids, beyond the crystalline-metal references [21-23]. If prior observations exist or are ambiguous, the claim should be reworded accordingly.","section":"Section III, Discussion; Introduction"}],"minor_comments":[{"comment":"The sentence 'beyond this point, the system transitions to the classical Bauschinger effect, where the yield stress in the reverse direction decreases, a phenomenon well-documented [].' contains an empty citation; please fill in the reference.","section":"Section II, Fig. 3 text"},{"comment":"The text refers to 'Fig. (a)' and 'Figs. (b-f)' without the figure number; please correct these to 'Fig. 3(a)' and 'Fig. 3(b-f)' for clarity.","section":"Section II, Fig. 3 caption and text"},{"comment":"Several mathematical symbols appear garbled in the text (for example, '¯Ã', 'Ãij', and the square-root radicals in Eq. (6)). Please ensure the final typeset version renders these correctly, as the current encoding makes the potential and fabric-tensor definitions hard to read.","section":"Section I, Models and Methods"},{"comment":"The caption states 'parameters τp = 1.0 and f=2.0'; the active force subscript is missing and should be 'f0 = 2.0' for consistency with the main text.","section":"Supplementary Information, Fig. 2 caption"},{"comment":"The supplementary text says 'Movie Files are added', but no files or links are listed. Please provide explicit file names or links so that reviewers and readers can access the claimed visual evidence.","section":"Supplementary Information, Movie Files"}],"recommendation":"major_revision","confidential_remarks":"The novelty claim of a 'first documented observation' should be checked carefully during editorial processing, because it is a strong claim and the supporting statistical evidence is currently thin. The manuscript also depends on companion work [19] that is unpublished and 'in preparation'; if that companion is central to the interpretation of the active shear-band network, the editor may wish to consider whether an updated version with more complete evidence is needed before further review."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a plausible simulation study of a genuinely new memory effect—an inverse Bauschinger response in active ultrastable glasses—but the central claim is not yet backed by the statistical or control evidence needed to believe it. The trend across pre-shear strain and persistence time is systematic, and the shear-band healing picture is attractive. I would not bet against the effect, but I also would not cite it as established.\n\nThe genuinely new thing is the observation that after pre-shearing an active ultrastable glass just past yield, the yield stress under reverse shear can exceed the forward yield stress. That is opposite to every amorphous-solid Bauschinger measurement I know. The crossover back to classical Bauschinger behavior at larger pre-strains and longer persistence times is a nice descriptor of where the effect lives. The authors also show, via the fabric tensor and non-affine displacement maps, that the anisotropy is localized along shear band networks. That is consistent and suggestive.\n\nThe main hole is a missing time control. The reverse test reaches yielding at a larger strain magnitude, so it runs for a longer simulated time before yielding. The active forces are on during that whole interval. The supplementary figure showing that stress stops evolving when activity is switched off demonstrates that the active state is time-dependent. It is therefore entirely possible that the 'healing' seen in Fig. 4(f) is simply activity-driven relaxation that would occur whether or not shear was reversed. A control in which the system is held at zero shear for the same elapsed time, then sheared forward, would separate directional memory from waiting-time aging. Without that control, the term 'healing upon shear reversal' is not justified by the data.\n\nThe second issue is statistics. All curves are single realizations. No system size is stated anywhere in the paper, and no error bars are given. In a system where stress-strain curves are famously sample-dependent, a factor-of-two difference in yield stress across two runs from the same state is not persuasive. The snapshots are nice but qualitative. The empty citation at 'well-documented []' on the classical Bauschinger effect is a sign of haste.\n\nNone of this means the effect is fake. The parameter dependence is coherent, and the mechanism—activity produces transient multiple shear bands that can be transiently healed by reversing the strain direction—is physically reasonable. It just needs to be tested properly. This paper is for people working on mechanical memory, yielding in glasses, and active matter; the question is interesting even if the evidence is incomplete.\n\nRecommendation: send it to a serious referee, but require the aging control, ensemble averaging, and a stated system size before publication. If those come back supporting the claim, this becomes a useful paper on mechanical memory in active glasses. As it stands, it is an unverified interesting observation.","headline":"Plausible and novel inverse Bauschinger claim in active glasses, but missing a time-matched aging control and any statistical averaging makes the central mechanism unproven.","tokens_in":16151,"tokens_out":4113,"would_cite":false,"duration_ms":42727,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Active glasses flip the Bauschinger memory effect","keywords":["inverse Bauschinger effect","ultrastable glasses","active matter","run-and-tumble particles","shear band networks","mechanical memory","Bauschinger effect","yielding"],"falsifier":"Repeat the shear-reversal protocol for $\\gamma_N=0.13$, $\\tau_p=0.1$, and $f_0=2.0$ on at least twenty independently prepared glass samples; if the mean reverse yield stress is not higher than the mean forward yield stress by more than the standard error, the inverse Bauschinger effect as reported is not statistically established.","tokens_in":15258,"feed_emoji":"🔄","tokens_out":8101,"duration_ms":75369,"temperature":0.7,"pith_summary":"The paper tries to show that an ultrastable glass—a deeply stabilized disordered solid that normally fails through one brittle shear band—can be made to remember the direction it was sheared in, but in the opposite sense from ordinary materials. In a standard Bauschinger effect, a material pre-sheared past yielding yields more easily when sheared backwards; here, active ultrastable glasses pre-sheared just past yielding yield at higher stress in the reverse direction than in the forward one. The cause, the authors argue, is that run-and-tumble active particles create a gradually forming network of shear bands that heals when the shear direction is reversed, temporarily erasing the weak spots. If true, the result would be the first documented inverse Bauschinger effect in an amorphous solid, and it would make shear history and activity a switch for writing, storing, or erasing mechanical memory in glasses.","feed_headline":"Active glasses flip the Bauschinger memory effect","feed_subtitle":"Run-and-tumble particles let shear bands heal, making reverse shear stronger, not weaker, than forward shear.","key_machinery":"The load-bearing object is the transient network of shear bands that forms when a fraction of particles is driven by persistent run-and-tumble forces during shear. In the two-dimensional polydisperse soft-sphere glass with $\\tau_p=0.1$ and $f_0=2.0$, activity turns the single brittle shear band of a passive ultrastable glass into a gradually evolving, branched network. The network is characterized by non-affine displacements $D^2_{\\min}$ and a fabric-tensor anisotropy index $\\alpha$ that measures the orientation bias of interparticle contacts; both show that plastic zones and contact anisotropy coincide. When shear is reversed shortly after yielding, this network dissolves at small reverse strains, so the system has to re-nucleate shear bands and yields later and at higher stress. Further pre-strain plasticizes the network, healing fails, and the classical Bauschinger effect returns.","core_discovery":"The authors claim that in a two-dimensional ultrastable glass doped with run-and-tumble active particles, reversing the shear direction at a strain just beyond yielding produces a yield stress on reverse loading that exceeds the forward yield stress—opposite to the classical Bauschinger effect. The effect appears at small pre-strains, for example $\\gamma_N=0.13$ with $\\tau_p=0.1$ and $f_0=2.0$, and disappears as the pre-strain grows: around $\\gamma_N\\approx0.26$ the forward and reverse responses become equal, and at larger strains the system returns to the classical Bauschinger response. Passive ultrastable glasses show only the classical effect. The paper connects the inverse effect to the transient healing of shear band networks: at the zero-stress state after unloading, active systems retain a network of mobile regions that anneals on reverse shear, delaying yield; at larger deformation the network becomes permanently plasticized and can no longer heal. The authors state their belief that this is the first documented inverse Bauschinger effect in amorphous solids.","pith_inferences":["The reported asymmetry rests on single trajectories; averaging over independent glass samples and reporting error bars would establish whether the inverse effect is a robust material property or a fluctuation.","A testable prediction follows directly: turning off activity before reversal should abolish the inverse effect and restore the passive drop in residual stress, which the supplementary activity-off runs already hint at.","The mechanism suggests that any glass that yields through a transient multi-band network—for instance passive glasses sheared at high rates—might show an inverse Bauschinger effect if the network can heal, so the phenomenon need not require activity itself.","If the effect is robust, active particles could serve as a reversible eraser for shear history: pre-shear writes a memory, reversal erases it, and repeated cycling overwrites it with a softer steady state."],"forward_implications":["A glass that fails through a single shear band cannot show the inverse effect; activity-induced multi-band networks are required, so active doping is a control parameter for the sign of the Bauschinger response.","There is a tunable crossover: small pre-strains (near $\\gamma_N=0.13$) give the inverse Bauschinger effect, while larger pre-strains give the classical Bauschinger effect, so deformation history alone selects the memory response.","Cyclic shear of active ultrastable glasses above the yield point causes progressive shear softening, with yield stress falling each cycle and a less-branched, permanent shear band network emerging before steady state.","Reversing shear can heal a just-formed shear band network, meaning a material can be made stronger against deformation in the opposite direction, not weaker."],"supporting_citations":[{"why":"Documents orientation-dependent and inverse Bauschinger behavior in crystalline solids, the contrast class for the amorphous claim.","marker":"[12]"},{"why":"Establishes the origin of the Bauschinger effect in amorphous solids, defining the classical asymmetry the active system is compared with.","marker":"[13]"},{"why":"Earlier work on plasticity-induced anisotropy and the Bauschinger effect in amorphous solids that provides the baseline for the yield-stress comparison.","marker":"[14]"},{"why":"Concurrent work showing active particles mitigate brittle failure in ultrastable glasses; the gradual shear-band formation described there underpins the healing mechanism.","marker":"[19]"},{"why":"Shows shear-band network structures can emerge in passive ultrastable glasses at high strain rates, giving the network-forming precedent.","marker":"[20]"},{"why":"Established strain localization and shear-band behaviour above the yield point under cyclic deformation, the setting the paper's healing result extends.","marker":"[24]"},{"why":"Shows amorphous silica develops permanent structural anisotropy under shear, motivating the fabric-tensor analysis of residual structure.","marker":"[26]"},{"why":"Supplies the polydisperse non-additive soft-sphere potential that defines the model ultrastable glass.","marker":"[33]"},{"why":"Supplies the swap Monte Carlo protocol used to prepare the deeply annealed, ultrastable samples.","marker":"[34]"},{"why":"Defines the $D^2_{\\min}$ non-affine displacement measure used to visualize shear bands and plastic zones.","marker":"[38]"}],"fun_headline_variants":["Active glasses invert Bauschinger yield","Shear memory flips Bauschinger in ultrastable glass","Run-and-tumble particles heal shear bands on reversal","First inverse Bauschinger effect in active glasses","Reverse shear yields stronger in active ultrastable glass"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim rests on the assumption that the higher reverse yield stress seen in Fig. 3(c) is a real property of the active glass and not the fluctuation of a single simulation run, since the stress-strain curves are shown without ensemble averaging, error bars, or a stated system size.","fun_headline_variants_meta":{"raw":{"variants":["Active glasses invert Bauschinger yield","Shear memory flips Bauschinger in ultrastable glass","Run-and-tumble particles heal shear bands on reversal","First inverse Bauschinger effect in active glasses","Reverse shear yields stronger in active ultrastable glass"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000342,"raw_usage":{"total_tokens":1877,"prompt_tokens":932,"completion_tokens":945,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":548,"completion_tokens_details":{"reasoning_tokens":870}},"tokens_in":548,"tokens_out":945,"duration_ms":9705,"temperature":1.0,"reasoning_tokens":870,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:17:59.573951+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the shear-reversal protocol for $\\gamma_N=0.13$, $\\tau_p=0.1$, and $f_0=2.0$ on at least twenty independently prepared glass samples; if the mean reverse yield stress is not higher than the mean forward yield stress by more than the standard error, the inverse Bauschinger effect as reported is not statistically established.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents orientation-dependent and inverse Bauschinger behavior in crystalline solids, the contrast class for the amorphous claim."},{"cited_title":"Patinet, A","cited_arxiv_id":null,"evidence_quote":"Establishes the origin of the Bauschinger effect in amorphous solids, defining the classical asymmetry the active system is compared with."},{"cited_title":"Karmakar, E","cited_arxiv_id":null,"evidence_quote":"Earlier work on plasticity-induced anisotropy and the Bauschinger effect in amorphous solids that provides the baseline for the yield-stress comparison."},{"cited_title":"Priya, J","cited_arxiv_id":null,"evidence_quote":"Concurrent work showing active particles mitigate brittle failure in ultrastable glasses; the gradual shear-band formation described there underpins the healing mechanism."},{"cited_title":"Singh, M","cited_arxiv_id":null,"evidence_quote":"Shows shear-band network structures can emerge in passive ultrastable glasses at high strain rates, giving the network-forming precedent."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Established strain localization and shear-band behaviour above the yield point under cyclic deformation, the setting the paper's healing result extends."},{"cited_title":"Similarly, our study also observes this behavior","cited_arxiv_id":null,"evidence_quote":"Shows amorphous silica develops permanent structural anisotropy under shear, motivating the fabric-tensor analysis of residual structure."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the swap Monte Carlo protocol used to prepare the deeply annealed, ultrastable samples."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the $D^2_{\\min}$ non-affine displacement measure used to visualize shear bands and plastic zones."}],"review_version":1}