Append-only AI overlay records; the canon claim file is never mutated.
Provisionally supportedconfidence: moderate · Claude Fable 5 (Cursor) · run orch-or-2026-08-22-fable-1 · unreviewed draft
Blinded, tolerance-controlled, large standardized effects, direction reproduced across species. Held at provisional by N = 8, males only, non-counterbalanced design, single-concentration latency rather than dose–response, and the fact that both studies share a senior author and program — a within-program replication, not independent confirmation.
Provisionally supportedconfidence: moderate · Gemini 3.1 Pro (Google) · run 2026-08-23-gemini31-1 · unreviewed draft
Blinded, tolerance-controlled studies in both rats and mice show a statistically significant delay in isoflurane-induced loss of righting reflex following acute epothilone B administration. Support is provisional because it relies on a single research program without independent confirmation. Furthermore, LORR is a behavioral proxy, and epothilone B alters numerous conventional cellular variables beyond quantum states.
Provisionally supportedconfidence: moderate · GPT-5.1 (OpenAI), independent judge run · run 2026-08-23-gpt51-1 · unreviewed draft
Khan and Huang, from the same group, report blinded LORR measurements showing acute epothilone B injections delay isoflurane-induced LORR by 69 s in rats (d≈1.9) and 29 s in mice (d≈0.8), with tolerance controls. Both studies are peer-reviewed and methodologically sound for small-N behavioral pharmacology, and directions are consistent across species. However, sample sizes are modest, the design is not a fully counterbalanced crossover, both studies share a senior author (so replication is not independent), and the effect is seen at a single high isoflurane concentration. Furthermore, Li et al. report opposite directional effects with related drugs under different regimens. The main objection is that, absent independent multicenter replication and full concentration–response curves, this result is best viewed as a promising but still tentative effect.
Provisionally supportedconfidence: moderate · Grok 4.6 (xAI), independent judge run · run 2026-08-23-grok46-1 · unreviewed draft
Both Wellesley experiments are peer-reviewed, used blinded LORR measurement, controlled for isoflurane tolerance, and found the same direction across species, with large standardized effects in the rat study (d = 1.9; +69 s) and a smaller one in mice (d = 0.8; +29 s). Against a stronger label: the rat analysis is N = 8, both studies used males only, dosing was not a counterbalanced crossover, both used latency at a single high concentration rather than a concentration–response curve, and they share a senior author, so this is within-program replication of direction. The strongest opposing consideration is Li et al. (ORCH-C024): an independent laboratory measuring the better-specified quantity found the opposite direction with a related stabilizer, which confines this result to a particular compound and regimen rather than a general microtubule-stability rule.
Provisionally supportedconfidence: moderate · Kimi K3 (Moonshot, via Venice) — independent check run via kimi-k3 · run 2026-08-23-check-kimi · unreviewed draft
Assessed as the behavioral claim it is: blinded LORR measurement, a tolerance control, large standardized effects (d = 1.9 rats, d = 0.8 mice), and direction reproduced across two species. It is held to provisional because the rat N is 8, both studies used males only, the design was not a counterbalanced crossover, and both studies share a senior author and research program — a within-program replication of direction, not independent confirmation, with the mouse paper itself conceding the effect is modest against inter-animal variability. The strongest opposing consideration is the independent contradictory result (ORCH-C024), which measured the better-specified quantity — full concentration–response curves — and found the opposite direction; latency at a single high concentration cannot adjudicate that conflict.
Provisionally supportedconfidence: moderate · Opus 5 (Claude), independent judge run · run 2026-08-23-opus5-1 · unreviewed draft
Both experiments are peer-reviewed, both used blinded loss-of-righting measurement, both controlled for tolerance from repeated isoflurane exposure, and the standardized effects are large to moderate with the direction reproducing across species. That is a real behavioral result and it is assessed as worded, independent of what it is offered to support. It does not reach well supported because the rat sample is N = 8, males only, the design was not a counterbalanced crossover, and the two studies share a senior author and research program — a within-program replication of direction, not independent confirmation. The authors themselves concede the 29 s mouse effect is not large against inter-animal variability. The strongest opposing consideration is that a latency shift at a single high concentration does not establish a shift in the concentration–response curve, which is exactly the quantity the independent study measured and found running the other way.
Provisionally supportedconfidence: moderate · anthropic/claude-opus-5 (scheduled maintenance run) · run 2026-08-25-auto-6bwy · unreviewed draft
Blinded, tolerance-controlled, large standardized effect in rats (ORCH-E011) with direction reproduced in mice, but ORCH-E012 is within-program and ORCH-E029 shows the 'independent confirmation' framing does not hold. Design limits (N=8, males only, non-counterbalanced, single isoflurane concentration) cap this at provisional.
Provisionally supportedconfidence: moderate · anthropic/claude-opus-5 (scheduled maintenance run) · run 2026-08-25-auto-jlsi · unreviewed draft
Blinded, tolerance-controlled, large standardized effect in rats with direction reproduced in mice, but ORCH-E012 is within-program and ORCH-E029 shows the independent-confirmation framing does not hold. Design limits (N=8, males, non-counterbalanced, single concentration) keep this at provisional.
Provisionally supportedconfidence: moderate · anthropic/claude-opus-5 (scheduled maintenance run) · run 2026-08-25-auto-scxb · unreviewed draft
Blinded with a tolerance control and a large standardized effect in rats, direction reproduced in mice — but ORCH-E012 is within-program rather than independent, and ORCH-E029 shows the 'independent confirmation' framing does not hold. Design limits (N=8, males, non-counterbalanced, single concentration) cap it here.
Provisionally supportedconfidence: moderate · Gemini 3.1 Pro (Google) — independent check run via gemini-3.1-pro-preview · run 2026-08-25-check-gemini · unreviewed draft
Two peer-reviewed, blinded studies from the same laboratory demonstrated that a microtubule stabilizer delays the isoflurane-induced loss of righting reflex in rats and mice, with large to moderate standardized effect sizes. However, they rely on latency at a single high concentration rather than full dose-response curves, lack independent outside replication, and manipulate general cellular integrity. The effect measures a behavioral proxy without recording any quantum variable.
Provisionally supportedconfidence: moderate · GPT-5.1 (OpenAI) — independent check run via gpt-5.1 · run 2026-08-25-check-gpt · unreviewed draft
The Wellesley studies show, with blinded LORR scoring and tolerance controls, that a single low-dose epothilone B injection produces statistically robust delays in isoflurane-induced LORR in both rats and mice, with large (rats) to moderate (mice) standardized effect sizes. These are straightforward behavioral findings and are internally consistent across species within the same program. The main limiting factors are small sample sizes, lack of counterbalanced crossover, shared authorship between the rat and mouse experiments, and the direct contradiction from an independent lab’s chronic epothilone D result, which shows that stabilizer effects are not uniformly protective.
Provisionally supportedconfidence: moderate · Grok 4.6 (xAI) — independent check run via grok-4.6 · run 2026-08-25-check-grok · unreviewed draft
Two peer-reviewed, blinded, tolerance-controlled studies from the same program report delayed isoflurane loss of righting reflex after acute epothilone B — 69 s in rats (d = 1.9) and 29 s in mice (d = 0.8) — which supports the behavioral effect as worded. Limits that keep the grade provisional include N = 8 males in the rat arm, no counterbalanced crossover, latency at a single high concentration rather than a concentration–response curve, and a within-program rather than independent replication. The strongest opposing consideration is the independent Li et al. concentration–response result running the opposite direction under a different compound and regimen, which does not erase this measurement but blocks any simple reading of it as "stabilization protects consciousness."
Provisionally supportedconfidence: moderate · Kimi K3 (Moonshot, via Venice) — independent check run via kimi-k3 · run 2026-08-25-check-kimi · unreviewed draft
Assessing only the behavioral effect as worded: both experiments are peer-reviewed, blinded, controlled for tolerance, and show large standardized effects (d = 1.9 rats; d = 0.8 mice) with the direction reproduced across species. It is held to provisional because the rat sample is N = 8, both studies used males only, the design was not a counterbalanced crossover, the latency was measured at a single high isoflurane concentration rather than a full concentration–response curve, and both studies share a senior author and program — a within-program replication of direction, not independent confirmation. The strongest opposing consideration is that LORR latency is a behavioral proxy whose endpoints fractionate across distinct receptors, and the independent fuller study (ORCH-C024) found the opposite direction with the better-specified measurement; epothilone B also perturbs far more about a neuron than any quantum state.
Provisionally supportedconfidence: moderate · Opus 5 (Anthropic) — independent check run via claude-opus-5 · run 2026-08-25-check-opus · unreviewed draft
Two peer-reviewed experiments with blinded LORR measurement and a tolerance control reproduce the direction across species with large standardized effects (d = 1.9 rats, d = 0.8 mice), and the claim as worded asserts only the behavioral delay. Held below "well_supported" because N = 8 in the main rat analysis, both studies used males only, the design was not a counterbalanced crossover (drug phase always followed the pre-drug phase), and the two studies share a senior author and programme — a within-programme replication of direction, not independent confirmation. The strongest opposing consideration is that a latency shift at a single high concentration (4% isoflurane) does not establish a shift in the concentration–response curve, and the independent study that did measure the curve found the opposite direction; the mice paper itself concedes the 29 s effect is small against inter-animal variability.
Provisionally supportedconfidence: moderate · anthropic/claude-opus-5 (scheduled maintenance run) · run 2026-08-26-auto-c256 · unreviewed draft
Blinded, tolerance-controlled, large standardized effect in rats with direction reproduced in mice — but ORCH-E012 is within-program rather than independent and ORCH-E029 shows the 'independent confirmation' framing does not hold. Design limits (N=8, males only, non-counterbalanced, single isoflurane concentration) cap it here.
Provisionally supportedconfidence: moderate · anthropic/claude-opus-5 (scheduled maintenance run) · run 2026-08-27-auto-mxfv · unreviewed draft
Blinded, tolerance-controlled, large standardized effect in rats with direction reproduced in mice — but ORCH-E012 is within-program rather than independent and ORCH-E029 shows the 'independent confirmation' framing does not hold. Design limits (N=8, males only, non-counterbalanced, single isoflurane concentration) cap it here.
Provisionally supportedconfidence: moderate · Gemini 3.1 Pro (Google) — independent check run via gemini-3.1-pro-preview · run 2026-08-27-check-gemini · unreviewed draft
Khan and Huang observed statistically significant delays in isoflurane-induced loss of righting reflex following acute epothilone B administration in rats and mice (ORCH-E011, ORCH-E012). The strongest opposing consideration is that these studies share a senior author, use latency at single concentrations rather than full dose-response curves, and rely on a behavioral proxy (LORR) that fractionates from consciousness.
Provisionally supportedconfidence: moderate · GPT-5.1 (OpenAI) — independent check run via gpt-5.1 · run 2026-08-27-check-gpt · unreviewed draft
Khan and Huang (E011–E012) report that acute epothilone B prolongs isoflurane LORR latency by 69 s in rats (d≈1.9) and 29 s in mice (d≈0.8) under blinded measurement with tolerance controls, directly supporting the stated latency shifts. However, both studies originate from the same group, with small N and without counterbalanced crossover or concentration–response curves, so this is within-program replication rather than independent confirmation. The strongest opposing consideration is the contradictory chronic epothilone-D data from an independent lab (ORCH-C024/E013), which show increased sensitivity rather than resistance and undermine generalization of the observed delay into a simple stabilizer rule.
Provisionally supportedconfidence: moderate · Grok 4.6 (xAI) — independent check run via grok-4.6 · run 2026-08-27-check-grok · unreviewed draft
Both Wellesley experiments are peer-reviewed, used blinded LORR measurement, controlled for isoflurane tolerance, and report large standardized effects (d = 1.9 in rats, d = 0.8 in mice) in the same direction across two species. Held to provisional by N = 8 males in the rat analysis, non-counterbalanced phase order, latency at a single high concentration rather than a concentration–response curve, shared senior author across the two papers, and the mice paper's own concession that 29 s is not large compared with inter-animal variability. Strongest opposing consideration: the independent Li et al. study measured the better-specified quantity — full LORR curves — and found the opposite direction with a different epothilone and regimen, and neither Wellesley experiment measured a microtubule quantum variable or an immobility endpoint.
Provisionally supportedconfidence: moderate · Kimi K3 (Moonshot, via Venice) — independent check run via kimi-k3 · run 2026-08-27-check-kimi · unreviewed draft
Both experiments are peer-reviewed, blinded, controlled for isoflurane tolerance, and report large standardized effects (d = 1.9 rats, d = 0.8 mice) with the direction reproduced across species. Provisional because the rat sample is N = 8, males only, every animal received the drug phase after the pre-drug phase rather than in counterbalanced order, and both studies share a senior author and program — a within-program replication of direction, not independent confirmation — with the mouse paper conceding the effect is modest against inter-animal variability. The strongest opposing consideration is that latency to LORR at a single high concentration is a behavioral proxy that does not locate a shift in the concentration–response curve, which is what the independent study measured and found running the other way (ORCH-E013); and epothilone B changes far more about a neuron than any quantum state. No microtubule variable of any kind was measured.
Provisionally supportedconfidence: moderate · Opus 5 (Anthropic) — independent check run via claude-opus-5 · run 2026-08-27-check-opus · unreviewed draft
Both experiments are peer-reviewed, blinded on the LORR endpoint, and controlled for isoflurane tolerance, with a large standardized effect in rats and the same direction in mice (ORCH-E011, ORCH-E012). Against a stronger grade: N = 8 in the main rat analysis, males only, no counterbalanced crossover, latency at a single high concentration rather than a concentration–response curve, and a shared senior author across both studies, so this is within-program replication of direction. The strongest opposing consideration is that an independent laboratory measuring the better-specified quantity found the opposite sign for a related compound (ORCH-E013); that does not falsify this measurement, since compound and regimen differ, but it means the effect should not yet be treated as a stable feature of microtubule pharmacology.
Provisionally supportedconfidence: moderate · anthropic/claude-opus-5 (scheduled maintenance run) · run 2026-09-01-auto-8fv7 · unreviewed draft
Blinded, tolerance-controlled, large standardized effect in rats with direction reproduced in mice — but ORCH-E012 is within-program rather than independent and ORCH-E029 shows the 'independent confirmation' framing does not hold. Design limits (N=8, males only, non-counterbalanced, single isoflurane concentration) cap it here.
Provisionally supportedconfidence: moderate · anthropic/claude-opus-5 (scheduled maintenance run) · run 2026-09-01-auto-ledp · unreviewed draft
Blinded, tolerance-controlled, large standardized effect in rats with direction reproduced in mice — but ORCH-E012 is within-program rather than independent and ORCH-E029 shows the 'independent confirmation' framing does not hold. Design limits (N=8, males only, non-counterbalanced, single isoflurane concentration) cap it here.
Provisionally supportedconfidence: moderate · Gemini 3.1 Pro (Google) — independent check run via gemini-3.1-pro-preview · run 2026-09-01-check-gemini · unreviewed draft
Blinded studies demonstrate that acute epothilone B delays LORR in rodents. The strongest opposing consideration is that epothilone B induces massive cellular, transport, and metabolic disruption, offering classical physiological routes to alter anesthetic latency without implicating any microtubule quantum state.
Provisionally supportedconfidence: high · GPT-5.1 (OpenAI) — independent check run via gpt-5.1 · run 2026-09-01-check-gpt · unreviewed draft
The Wellesley rat and mouse studies use blinded LORR measurement, show statistically significant delays (69 s, d=1.9 in rats; 29 s, d=0.8 in mice) after acute epothilone B, and control for tolerance, which supports the claim that acute stabilizer dosing delays LORR under isoflurane. However, sample sizes are small, both studies come from the same group, dosing order is not counterbalanced, and the mice effect is modest amid high variability. The strongest opposing consideration is that independent chronic epothilone D data shift sensitivity in the opposite direction (ORCH-C024), and no microtubule‐specific or quantum variable was measured, leaving open many conventional explanations.
Provisionally supportedconfidence: moderate · Grok 4.6 (xAI) — independent check run via grok-4.6 · run 2026-09-01-check-grok · unreviewed draft
Khan et al. report that a single low dose of epothilone B delayed isoflurane LORR by 69 s in rats (Cohen's d = 1.9), blinded, with a tolerance control; Huang et al. reproduced the direction in mice (+29 s, d = 0.8). Assessed strictly as a behavioral effect, that is a real within-program result. The strongest opposing consideration is that both studies share a senior author and program, the rat analysis is N = 8 males only without a counterbalanced crossover, latency at one high concentration is not a concentration–response curve, and Li et al. 2025 measured full curves with a related stabilizer and found the opposite direction. No microtubule quantum variable was measured in either Wellesley experiment.
Provisionally supportedconfidence: moderate · Kimi K3 (Moonshot, via Venice) — independent check run via kimi-k3 · run 2026-09-01-check-kimi · unreviewed draft
Assessed as the behavioral effect it claims: two peer-reviewed studies, blinded LORR measurement, a tolerance control, large standardized effects (d = 1.9 rats, d = 0.8 mice), and the direction reproduced across species. The provisional grade reflects N = 8 in the rat analysis, males only, non-counterbalanced phase ordering, and — most importantly — both studies sharing a senior author and research program, making this within-program replication rather than independent confirmation; the mouse paper itself concedes the effect is modest against inter-animal variability. The strongest opposing consideration is that latency at a single high concentration is not a shift in the concentration–response curve, and the independent study that measured the full curve found the opposite direction (ORCH-E013) — while LORR is a behavioral proxy that fractionates from other anesthetic endpoints, and no microtubule quantum variable was measured.
Provisionally supportedconfidence: moderate · Opus 5 (Anthropic) — independent check run via claude-opus-5 · run 2026-09-01-check-opus · unreviewed draft
Both experiments are peer-reviewed with blinded LORR measurement, a tolerance control, and large standardized effects (d = 1.9 rats, d = 0.8 mice), and the direction reproduced across species — that is real support for the behavioural effect as worded. It cannot rise higher because N = 8 in the main rat analysis, males only, no counterbalanced crossover, latency at a single high concentration rather than a concentration–response curve, and both studies share a senior author and program, so this is within-program replication of direction. The strongest opposing consideration is that the independent study measuring the better-specified quantity (ORCH-E013) found sensitivity shifting the other way, and the authors themselves concede the mouse effect is small against inter-animal variability.
Provisionally supportedconfidence: moderate · anthropic/claude-opus-5 (scheduled maintenance run) · run 2026-09-02-auto-79ed · unreviewed draft
Blinded, tolerance-controlled, large standardized effect in rats with the direction reproduced in mice — but ORCH-E012 is within-program rather than independent and ORCH-E029 shows the 'independent confirmation' framing does not hold. Design limits (N=8, males only, non-counterbalanced, single isoflurane concentration) cap it here.
Provisionally supportedconfidence: moderate · anthropic/claude-opus-5 (scheduled maintenance run) · run 2026-09-02-auto-tyck · unreviewed draft
Blinded, tolerance-controlled, large standardized effect in rats with direction reproduced in mice — but ORCH-E012 is within-program rather than independent and ORCH-E029 shows the 'independent confirmation' framing does not hold. Design limits (N=8, males only, non-counterbalanced, single isoflurane concentration) cap it here.
Provisionally supportedconfidence: moderate · none — mechanical migration (scripts/migrations/2026-09-08-editions.ts) · run 2026-09-08-migration-orch-or · unreviewed draft
Evidence-only, assessing just the behavioral effect as worded. Both experiments are peer-reviewed, both used blinded LORR measurement, both controlled for tolerance from repeated isoflurane exposure, and the effect sizes are large in standardized terms (d = 1.9 rats; d = 0.8 mice). The direction reproduced across species. Against a stronger label: the rat sample is N = 8, both studies used males only, animals received the drug phase after the pre-drug phase rather than in a counterbalanced crossover, and the two studies share a senior author and research program — so this is a within-program replication of direction, not independent confirmation. The mice paper concedes the 29 s effect "is not large compared to the variability among different animals."
Provisionally supportedconfidence: moderate · claude-opus-5 · run 2026-09-11-edition-124802 · unreviewed draft
ORCH-E011 is the cleanest behavioural record on the proponent side: blinded, large standardized effect (d = 1.9), with a control against anesthesia tolerance. ORCH-E012 reproduces the direction across species at 29 s and d = 0.8. Two things hold the verdict here rather than higher. The mouse study is the same laboratory and senior author, so it is a replication of direction within one programme, and its own abstract concedes the effect is not large against inter-animal variability; ORCH-E029 records that the public claim of independent confirmation does not hold. And the rat design is N = 8, male only, latency at a single high isoflurane concentration, and not counterbalanced. The observation stands as reported; what it means is a separate matter.
Well supportedconfidence: high · Gemini 3.8 Flash (Google) — independent check run via gemini-3.8-flash · run 2026-09-30-check-gemini-104603 · unreviewed draft
Blinded rodent behavioral experiments by Khan et al. demonstrated that acute epothilone B delayed isoflurane-induced loss of righting reflex by 69 seconds in rats (Cohen's d = 1.9), and Huang et al. extended this with a 29-second delay in mice (d = 0.8). Both studies included appropriate controls against anesthetic tolerance. The strongest opposing consideration is that both studies originate from the same laboratory and research program, utilizing modest sample sizes without counterbalanced crossover allocations.
Provisionally supportedconfidence: moderate · GLM 5.3 Flash (Z.ai, via Venice) — independent check run via z-ai-glm-5-3-flash · run 2026-09-30-check-glm-104603 · unreviewed draft
Both effects are verified in peer-reviewed abstracts, blinded, with a tolerance control and a large standardized effect in rats. But the mouse replication is within-program with a shared senior author, not independent, and the effect shrank and decayed; no quantum variable was measured; and Li et al. 2025's contrary independent result constrains how the finding generalizes even though different compound and regimen prevent direct contradiction. Provisional pending the multicenter replication the agenda correctly proposes.
Well supportedconfidence: moderate · GPT-5.6 Sol (OpenAI) — independent check run via gpt-5.6-sol · run 2026-09-30-check-gpt-104603 · unreviewed draft
ORCH-E011 directly reports a 69-second rat delay after acute epothilone B, and ORCH-E012 reports a 29-second mouse delay in the same research program. The exact directional observation is therefore reproduced across two species, although not independently across laboratories. The strongest opposing consideration is the small rat sample and the fact that the independent chronic epothilone-D study found increased rather than decreased anesthetic sensitivity.
Well supportedconfidence: high · Grok 4.5 (xAI) — independent check run via grok-4.5 · run 2026-09-30-check-grok-104603 · unreviewed draft
Khan et al. report a blinded +69 s LORR delay after acute low-dose epothilone B in rats with large effect size and a tolerance control; Huang et al. extend directionally to mice (+29 s) in the same program. Strongest opposing consideration is small N, non-counterbalanced design, single high isoflurane concentration, no quantum variable, and non-independence of the mouse replication.
Provisionally supportedconfidence: moderate · claude-opus-5-5 · run 2026-09-30-edition-094411 · unreviewed draft
ORCH-E011 (rats, blinded, tolerance control, d = 1.9) and ORCH-E012 (mice, d = 0.8, same laboratory) support the stated latencies. ORCH-E029 records that the two studies are not independent. The rat study has N = 8, males only, and no counterbalanced crossover. The observation is credible; what it means is not established.