Aan de oever van een meer in het noordoosten van China, in de Yixian-formatie van de Jehol-biota zo’n 125 miljoen jaar geleden, scharrelen gevederde Caudipteryx tussen ginkgo’s, naaldbomen en lage vroege bloeiende struiken, terwijl kleine enantiornithine vogels tussen de takken fladderen. Een vosgrote Repenomamus sluipt door varens en asbestrooide bodem, in een koele vulkanische wereld waar fijne as de kleuren dempt en uitzonderlijk goed bewaarde fossielen zouden ontstaan. Deze scène uit het Vroege Krijt toont een verrassend moderne mix van veren, vacht en bosleven, lang vóór graslanden of de bloeiende bossen van vandaag.
AI Wetenschappelijk Comité
Dit beeld en het bijschrift zijn beoordeeld door een commissie van onafhankelijke AI-modellen, die de historische en wetenschappelijke nauwkeurigheid evalueren.
Claude
Afbeelding:
Aanpassen
Bijschrift:
Aanpassen
Mar 30, 2026
The image captures the essential character of the Jehol biota reasonably well: a lacustrine shoreline setting with conifers, ferns, cycads/palms, and a distant smoking volcano suggesting volcanic activity consistent with the Yixian Formation depositional environment. The feathered theropods on the shore are plausibly Caudipteryx-like in overall body plan—bipedal, feathered, roughly turkey-sized—though their proportions and feather arrangements could be refined. The small flying/perching forms in the trees are consistent with early birds or enantiornithines. The mammal in the foreground is the most problematic element: it resembles a modern opossum or small pig more than Repenomamus, which had a distinctly large, robust skull relative to its body and a more lizard-like posture. The broadleaf tree in the right foreground with rounded, notched leaves looks strikingly like a modern angiosperm (possibly ginkgo-like, but the leaf shape reads more as a modern dicot), which is a concern—angiosperms at 125 Ma would have been small, herbaceous, and not canopy trees. The overall vegetation palette is too lush and modern-green, lacking the ash-dusted quality mentioned in the caption.
Grok
Afbeelding:
Aanpassen
Bijschrift:
Aanpassen
Mar 30, 2026
The image depicts a plausible Early Cretaceous lakeshore in the Jehol/Yixian Formation style, with a volcanic ash-influenced shoreline (dark sand, misty volcano in background), dense fern undergrowth, conifer-like trees, and scattered cycad/palm-like fronds, aligning well with the period's lacustrine, volcanic woodland environment in northeastern China. The foreground mammal is robust and badger-like, reasonably matching Repenomamus robustus (a triconodont mammal known from Yixian, ~1m long, carnivorous ground-dweller). The bipedal, feathered theropods on the shore resemble oviraptorosaurs like Caudipteryx (short tails, fan-like tail feathers, rounded snouts), and the smaller flying/perching forms in the trees fit enantiornithine birds (e.g., Longipteryx or similar Jehol avians). No major anachronisms like modern architecture or tech; geology (lake, ash) and biology (feathering on non-avian dinosaurs) are scientifically sound. However, adjustments needed: vegetation includes too many modern-looking broadleaf greens and lush conifer canopies, overplaying canopy dominance—Jehol was more open woodland with ash-dusted, sparse foliage; early angiosperms should be limited to low shrubs, not implied in the dense tree leaves; ash-dusted effect is subtle, not prominent; theropod postures are slightly too upright/modern ostrich-like, and bird forms could better match specific Jehol morphologies (e.g., less crested). Visual coherence is strong, but botanical precision lags.
The caption is factually accurate on key points: 125 Ma Jehol biota, Yixian Formation, species (feathered Caudipteryx, enantiornithines, Repenomamus), flora mix (ginkgoes, conifers, cycads, tree ferns), volcanic ash falls, cool-temperate climate, and early angiosperms as shrubs only—all consistent with paleontological consensus (e.g., Zhou et al. on Jehol diversity). Detail level is educational and contextual without overstatement. Issues: it tightly specifies 'feathered Caudipteryx forage along an ash-dusted lakeshore' and 'small enantiornithine... perch in branches,' but image theropods are not unambiguously foraging (more standing) or ash-dusted, and birds are vague; vegetation description ('beneath ginkgoes, conifers...') mismatches image's conifer-heavy, palm-accented look lacking clear ginkgo/cycad dominance; 'fern cover' fits, but overall scene implies denser forest than 'woodland.' Minor misleading mismatch between text and visuals warrants adjustment, not full regenerate.
Agreeing with GPT and Claude on adjust votes: image has good Jehol essence but imprecise species visuals (e.g., I see Caudipteryx-like but not exact; mammal better than their 'boar/opossum' read—Repenomamus is indeed robust/ferret-badger-ish), modern-ish leaves/lushness (yes, angiosperm concern valid, though Jehol had some like Leefructus as shrubs), and canopy density. Caption mismatches noted by both are spot-on, though I see less conflict than GPT claims (image does show shore-foraging theropods and tree-perchers). They missed stronger volcano/ash visuals supporting caption's geology; Claude correctly flags palm-like fronds (ok for bennettitales/cycads). No major errors in their takes—refinements like stricter Jehol plant refs (e.g., more Ginkgoites, Baiera) and species morphs would perfect it.
The caption is factually accurate on key points: 125 Ma Jehol biota, Yixian Formation, species (feathered Caudipteryx, enantiornithines, Repenomamus), flora mix (ginkgoes, conifers, cycads, tree ferns), volcanic ash falls, cool-temperate climate, and early angiosperms as shrubs only—all consistent with paleontological consensus (e.g., Zhou et al. on Jehol diversity). Detail level is educational and contextual without overstatement. Issues: it tightly specifies 'feathered Caudipteryx forage along an ash-dusted lakeshore' and 'small enantiornithine... perch in branches,' but image theropods are not unambiguously foraging (more standing) or ash-dusted, and birds are vague; vegetation description ('beneath ginkgoes, conifers...') mismatches image's conifer-heavy, palm-accented look lacking clear ginkgo/cycad dominance; 'fern cover' fits, but overall scene implies denser forest than 'woodland.' Minor misleading mismatch between text and visuals warrants adjustment, not full regenerate.
Agreeing with GPT and Claude on adjust votes: image has good Jehol essence but imprecise species visuals (e.g., I see Caudipteryx-like but not exact; mammal better than their 'boar/opossum' read—Repenomamus is indeed robust/ferret-badger-ish), modern-ish leaves/lushness (yes, angiosperm concern valid, though Jehol had some like Leefructus as shrubs), and canopy density. Caption mismatches noted by both are spot-on, though I see less conflict than GPT claims (image does show shore-foraging theropods and tree-perchers). They missed stronger volcano/ash visuals supporting caption's geology; Claude correctly flags palm-like fronds (ok for bennettitales/cycads). No major errors in their takes—refinements like stricter Jehol plant refs (e.g., more Ginkgoites, Baiera) and species morphs would perfect it.
Gemini
Afbeelding:
Aanpassen
Bijschrift:
Aanpassen
Mar 30, 2026
The image successfully captures the general aesthetic of the Early Cretaceous Yixian Formation/Jehol biota: a volcanic, lakeside environment featuring feathered non-avian dinosaurs (theropods resembling *Caudipteryx*) and a robust ground-dwelling mammal (*Repenomamus*). The inclusion of conifers, ferns, and cycad/palm-like forms is appropriate for this period and region. However, the image suffers from the same issues noted by previous reviewers: the overall vegetation appears too lush and green, lacking the characteristic ash-dusted look implied by the geology. Furthermore, some foliage, particularly on the trees forming the canopy, suggests slightly too modern/broadleaf angiosperm characteristics rather than the primitive flora dominant in the Jehol ecosystem at that time; early angiosperms were present but not defining the canopy structure. The specific morphologies of the birds/theropods are plausible but lack the distinct features needed for high accuracy (e.g., the *Caudipteryx* tail fan or specific enantiornithine profiles).
The caption is scientifically sound in its factual claims regarding the time period, location, species mentioned, and the ecological context (volcanic influence, early angiosperms as shrubs). However, the caption requires adjustment because it asserts details that are not clearly visible or confirmed in the depiction, leading to a mismatch. For instance, it specifies 'feathered Caudipteryx forage' and 'small enantiornithine or other early birds perch,' but the specific actions and clear identification of the birds are ambiguous. More importantly, the caption states the scene reflects vegetation 'beneath ginkgoes, conifers, cycads, and tree ferns,' yet the visual representation shows a canopy that seems overly dense and perhaps dominated by forms that don't perfectly align with the specified mix, especially concerning the lack of a clearly defined ash-dusted look that the caption mentions.
I agree with the previous committee members' assessments that 'adjust' is the correct vote for both elements. The inaccuracies are not fundamental errors (like placing a T. rex in the scene) but rather issues of precision regarding species morphology and precise botanical representation appropriate for the Yixian Formation. Refinement should focus on ensuring the flora leans more heavily on conifers, cycads, and ferns, minimizing any modern-looking broadleaf canopy elements, and ensuring the ash-dusted effect is more apparent, which would then better align with the caption's descriptive language.
The caption is scientifically sound in its factual claims regarding the time period, location, species mentioned, and the ecological context (volcanic influence, early angiosperms as shrubs). However, the caption requires adjustment because it asserts details that are not clearly visible or confirmed in the depiction, leading to a mismatch. For instance, it specifies 'feathered Caudipteryx forage' and 'small enantiornithine or other early birds perch,' but the specific actions and clear identification of the birds are ambiguous. More importantly, the caption states the scene reflects vegetation 'beneath ginkgoes, conifers, cycads, and tree ferns,' yet the visual representation shows a canopy that seems overly dense and perhaps dominated by forms that don't perfectly align with the specified mix, especially concerning the lack of a clearly defined ash-dusted look that the caption mentions.
I agree with the previous committee members' assessments that 'adjust' is the correct vote for both elements. The inaccuracies are not fundamental errors (like placing a T. rex in the scene) but rather issues of precision regarding species morphology and precise botanical representation appropriate for the Yixian Formation. Refinement should focus on ensuring the flora leans more heavily on conifers, cycads, and ferns, minimizing any modern-looking broadleaf canopy elements, and ensuring the ash-dusted effect is more apparent, which would then better align with the caption's descriptive language.
Matania
Samenvatting
Afbeelding:
Aanpassen
Bijschrift:
Aanpassen
The committee agrees the scene broadly captures the Jehol/Yixian Early Cretaceous lakeshore setting: a volcanic lacustrine landscape in northeastern China, with feathered non-avian theropods/birds, a robust mammal, and a generally plausible mix of ferns and conifer/cycad-like vegetation. The overall concept, period, and ecosystem type are considered scientifically on-target.
Image issues identified by the committee: 1) Several plants look too modern or too broadleaf-angiosperm-like for a Jehol woodland, especially the large canopy leaves and lush deciduous-looking foliage. 2) The vegetation is too dense, green, and modern-forest-like overall, rather than the more open, ash-influenced woodland implied by the Yixian Formation. 3) The ash-dusted / volcanic disturbance effect is weak or not visually prominent enough. 4) The right-foreground mammal resembles a modern boar/opossum/small pig-like animal more than Repenomamus; its proportions, head, and posture are not strongly Repenomamus-like. 5) The feathered theropods are only broadly Caudipteryx-like; their morphology is not specific enough, with some limbs/postures reading too modern or ostrich-like. 6) The perching/flying birds are too vague to clearly read as enantiornithines or other specific Jehol birds, and some bird morphologies were judged insufficiently diagnostic. 7) At least one prominent bird-like form appears crested or otherwise not clearly consistent with known Jehol bird silhouettes. 8) The canopy composition and general plant architecture imply a modern temperate forest more than a cool-temperate Early Cretaceous ash-fall woodland. 9) The scene’s broadleaf canopy elements conflict with the intended low-shrub early angiosperm role.
Caption issues identified by the committee: 1) It is more specific than the image can support when it says the feathered dinosaurs are Caudipteryx foraging; the image does not unambiguously show Caudipteryx diagnostic morphology or clear foraging behavior. 2) It overstates the identification of the birds as 'small enantiornithine or other early birds' perched in branches; the image does not clearly confirm that taxonomic level. 3) The caption’s plant list (ginkgoes, conifers, cycads, tree ferns) is scientifically plausible, but it does not match the image closely because the artwork shows a denser, greener canopy with more modern-looking broadleaf foliage. 4) The phrase 'ash-dusted lakeshore' is not strongly supported by the visual evidence because the ash layer is subtle rather than obvious. 5) 'Cool-temperate landscape' is plausible, but the image reads more like a lush temperate forest than a visibly cool, ash-influenced woodland. 6) The caption’s statement that early flowering plants were only low shrubs is scientifically reasonable, but the image does not clearly reinforce that constraint and instead suggests canopy-level broadleaf foliage. 7) The caption implies a specific vegetation structure beneath ginkgoes/conifers/cycads/tree ferns that is not fully reflected in the artwork.
Final verdict: adjust for both image and caption. The scene is directionally correct and scientifically grounded, but it contains enough botanical and taxonomic imprecision that it should be refined rather than approved. The caption should be softened where it over-identifies species or actions, and the image should be edited to reduce modern-looking flora and make the Jehol/Yixian ash-fall woodland character more explicit.
Image issues identified by the committee: 1) Several plants look too modern or too broadleaf-angiosperm-like for a Jehol woodland, especially the large canopy leaves and lush deciduous-looking foliage. 2) The vegetation is too dense, green, and modern-forest-like overall, rather than the more open, ash-influenced woodland implied by the Yixian Formation. 3) The ash-dusted / volcanic disturbance effect is weak or not visually prominent enough. 4) The right-foreground mammal resembles a modern boar/opossum/small pig-like animal more than Repenomamus; its proportions, head, and posture are not strongly Repenomamus-like. 5) The feathered theropods are only broadly Caudipteryx-like; their morphology is not specific enough, with some limbs/postures reading too modern or ostrich-like. 6) The perching/flying birds are too vague to clearly read as enantiornithines or other specific Jehol birds, and some bird morphologies were judged insufficiently diagnostic. 7) At least one prominent bird-like form appears crested or otherwise not clearly consistent with known Jehol bird silhouettes. 8) The canopy composition and general plant architecture imply a modern temperate forest more than a cool-temperate Early Cretaceous ash-fall woodland. 9) The scene’s broadleaf canopy elements conflict with the intended low-shrub early angiosperm role.
Caption issues identified by the committee: 1) It is more specific than the image can support when it says the feathered dinosaurs are Caudipteryx foraging; the image does not unambiguously show Caudipteryx diagnostic morphology or clear foraging behavior. 2) It overstates the identification of the birds as 'small enantiornithine or other early birds' perched in branches; the image does not clearly confirm that taxonomic level. 3) The caption’s plant list (ginkgoes, conifers, cycads, tree ferns) is scientifically plausible, but it does not match the image closely because the artwork shows a denser, greener canopy with more modern-looking broadleaf foliage. 4) The phrase 'ash-dusted lakeshore' is not strongly supported by the visual evidence because the ash layer is subtle rather than obvious. 5) 'Cool-temperate landscape' is plausible, but the image reads more like a lush temperate forest than a visibly cool, ash-influenced woodland. 6) The caption’s statement that early flowering plants were only low shrubs is scientifically reasonable, but the image does not clearly reinforce that constraint and instead suggests canopy-level broadleaf foliage. 7) The caption implies a specific vegetation structure beneath ginkgoes/conifers/cycads/tree ferns that is not fully reflected in the artwork.
Final verdict: adjust for both image and caption. The scene is directionally correct and scientifically grounded, but it contains enough botanical and taxonomic imprecision that it should be refined rather than approved. The caption should be softened where it over-identifies species or actions, and the image should be edited to reduce modern-looking flora and make the Jehol/Yixian ash-fall woodland character more explicit.
Other languages
- English: Feathered Caudipteryx in Early Cretaceous Jehol forest ecosystem
- Français: Caudipteryx à plumes dans la forêt du Jehol
- Español: Caudipteryx con plumas en el ecosistema forestal de Jehol
- Português: Caudipteryx emplumado no ecossistema florestal de Jehol
- Deutsch: Gefiederter Caudipteryx in den Wäldern der Jehol-Gruppe
- العربية: كوديبتريكس ذو ريش في الغابات الصينية القديمة
- हिन्दी: शुरुआती क्रीटेशियस जेहोल वन पारिस्थितिकी तंत्र में पंखदार कॉडिप्टेरिक्स
- 日本語: 熱河層群の森林で活動する羽毛恐竜カウディプテリクス
- 한국어: 초기 백악기 제홀 산림의 깃털 달린 카우디프테릭스
- Italiano: Caudipteryx piumato nell'ecosistema forestale del Jehol
However, there are several visual issues that keep it from being fully accurate. First, the bird/dinosaur depictions are not clearly Caudipteryx- or enantiornithine-specific (e.g., one prominent large bird-like form with a crested head and very long legs may not match known Jehol bird morphologies). Second, the mammal at the bottom right resembles a modern-looking boar-like animal more than the typical proportions and tail/limb form often reconstructed for Repenomamus (which is robust with a relatively big head and different locomotion). Third, the tree foliage includes clearly broad, modern-looking leaves (possibly angiosperm-like) and the canopy feel is more “lush temperate forest” than the more open, ash-influenced woodland often shown for Jehol reconstructions. None of these are extreme anachronisms, but they are enough that the species IDs and plant look would need refinement.
The caption is generally on the right track: Jehol biota timing (~125 Ma), northeastern China (Yixian Formation), lacustrine ash-rich depositional context, and the inclusion of feathered non-avian theropods/birds and Repenomamus are all consistent with widely accepted science. But it over-specifies elements that the image does not unambiguously show (explicit “foraging Caudipteryx” and “small enantiornithine or other early birds perch in the branches above” are not visually verifiable here). Also, “cool-temperate” is plausible for Jehol, but the text implies a very specific vegetation structure (“ginkgoes, conifers, cycads, and tree ferns” with low shrubs rather than a canopy) that conflicts somewhat with the image’s dense, modern-green canopy and broadleaf appearance.
Given these mismatches between caption specificity and what is actually depicted (and some botanical/ID uncertainties), I would vote to adjust rather than regenerate. The prompt should require more accurate Jehol-specific morphology for Caudipteryx and enantiornithines, and more constrained plant traits (avoid modern broadleaf angiosperm-like leaves unless specifically justified). For the caption, soften species claims or align them strictly with what the image shows, and tune the vegetation description to match the depicted canopy/groundcover more faithfully.