{"id":"b0af3bb1-4773-4373-99b2-cb438c5e6930","arxiv_id":"2605.30819","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Function2Scene is a framework that parses functional design briefs into a 17-criteria taxonomy of constraints and applies iterative geometric-LLM-VLM refinement to produce 3D layouts preferred over LLM baselines in 94.3% of comparisons on 30 cases.","lead":"Function2Scene generates 3D indoor scene layouts from natural-language briefs describing room users and their activities, deriving constraints from a 17-criteria taxonomy and refining via geometric, LLM, and VLM checks. A smart generalist might read it to see how AI scene generation can shift from object lists to supporting actual human use.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"No significant objection identified","rationale":"Reader correctly flagged inability to assess due to missing full text; with only abstract available here, no load-bearing technical flaw can be located in the argument as stated.","tokens_in":1754,"tokens_out":198,"duration_ms":14531,"concrete_test":"Extract the exact evaluation protocol from the results section (including number of raters, blinding procedure, and how baselines were prompted with functional specifications) and recompute the preference percentage after removing any cases where baseline prompts were not functionally equivalent.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The provided abstract and strongest claim contain no internally inconsistent assumptions or unsupported steps in the central experimental result. The method description (parsing to 17-criteria taxonomy followed by iterative check-and-repair) is presented at a level that does not reveal a logical gap that would invalidate the 94.3% preference claim on the 30 cases.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper introduces Function2Scene, a framework that generates 3D indoor scene layouts from natural-language functional specifications (design briefs describing occupants and their activities). It parses personas and activities, derives constraints from a fixed taxonomy of 17 criteria (spatial, ergonomic, activity, environmental), and performs iterative check-and-repair using geometric measurements, LLM contextual reasoning, and VLM visual assessment. On 30 professionally written cases, layouts are preferred over recent LLM-based baselines in 94.3% of pairwise comparisons, reframing scene synthesis around functional support rather than object placement.","tokens_in":1828,"tokens_out":556,"duration_ms":12202,"significance":"If the experimental result holds under rigorous controls, the work offers a substantive shift in text-driven 3D scene synthesis by prioritizing human-use constraints over plausibility. The taxonomy-driven constraint derivation and tool-augmented refinement loop are concrete technical contributions that could influence downstream applications in design software and embodied AI; the absence of free parameters or ad-hoc axioms in the core pipeline is a strength.","major_comments":[{"comment":"Experiments section: the 94.3% preference rate on 30 cases is the central empirical claim, yet the manuscript provides no details on baseline implementations, exact participant instructions, randomization of presentation order, or statistical significance testing; without these, the result cannot be assessed for robustness or selection effects.","section":"Experiments"},{"comment":"Method, refinement loop description: the iterative check-and-repair process combines LLM-based reasoning and VLM-based assessment in both generation and judgment steps; this raises a concrete risk of self-referential evaluation that is not addressed by any ablation or external validation, directly affecting the reliability of the functional-satisfaction claims.","section":"Method"},{"comment":"Method, taxonomy derivation: the claim that a fixed 17-criteria taxonomy can be reliably parsed from arbitrary natural-language briefs and then used to guide effective layout generation lacks supporting evidence (e.g., inter-annotator agreement or failure-case analysis) and is load-bearing for the entire pipeline.","section":"Method"}],"minor_comments":[{"comment":"Abstract and introduction: the baselines are referred to only as 'recent LLM-based scene synthesis baselines' without naming the specific methods or citations; this should be explicit for reproducibility.","section":"Abstract"},{"comment":"Notation: the distinction between 'functional specifications' and the derived 'constraints' is used interchangeably in places; consistent terminology would improve clarity.","section":null}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the thoughtful and constructive report. We address each major comment below, indicating where revisions to the manuscript are planned.","responses":[{"response":"We agree that these details are required to allow proper assessment of the result. The revised manuscript will expand the Experiments section with: complete specifications and implementation details for all baselines; the exact participant instructions and study protocol; confirmation that presentation order was randomized; the number of participants; and statistical significance testing (including p-values from a binomial or sign test on the preference data).","revision_made":"yes","referee_comment":"[Experiments] Experiments section: the 94.3% preference rate on 30 cases is the central empirical claim, yet the manuscript provides no details on baseline implementations, exact participant instructions, randomization of presentation order, or statistical significance testing; without these, the result cannot be assessed for robustness or selection effects."},{"response":"We acknowledge the risk of circularity. Geometric measurements provide an objective, non-LLM/VLM component. In the revision we will add an ablation that isolates each module (geometric only, geometric+LLM, full loop) and, on a subset of cases, report agreement with independent human judgments of functional satisfaction. We will also clarify the separation between generation and judgment prompts.","revision_made":"yes","referee_comment":"[Method] Method, refinement loop description: the iterative check-and-repair process combines LLM-based reasoning and VLM-based assessment in both generation and judgment steps; this raises a concrete risk of self-referential evaluation that is not addressed by any ablation or external validation, directly affecting the reliability of the functional-satisfaction claims."},{"response":"The taxonomy is drawn from established interior-design literature. The current manuscript does not contain an inter-annotator agreement study. The revision will include the complete taxonomy with definitions and few-shot examples, plus a failure-case analysis of parsing outcomes on the 30 evaluation briefs. This will provide concrete evidence of parsing behavior even if a full IAA study is not added.","revision_made":"partial","referee_comment":"[Method] Method, taxonomy derivation: the claim that a fixed 17-criteria taxonomy can be reliably parsed from arbitrary natural-language briefs and then used to guide effective layout generation lacks supporting evidence (e.g., inter-annotator agreement or failure-case analysis) and is load-bearing for the entire pipeline."}],"tokens_in":1474,"tokens_out":527,"duration_ms":20168,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The paper's main move is to take natural-language design briefs about occupants and activities, parse them into a fixed 17-criteria taxonomy, and then run an iterative check-and-repair loop that mixes geometric measurements, LLM reasoning, and VLM visual checks instead of asking an LLM to output a final layout in one shot.\n\nThat framing is new relative to the object-centric baselines mentioned. Treating layout as support for human use rather than plausible object placement is a useful shift for downstream uses in design and simulation. The decision to keep the LLM from doing everything at once and to add explicit geometric and visual tools is a practical engineering choice that avoids some obvious failure modes.\n\nThe reported result is a 94.3% preference rate over recent LLM baselines on 30 professionally written cases. The number is high, but the abstract supplies almost no information on baseline implementations, participant instructions, or how cases were selected. That makes it hard to judge whether the gap is real or an artifact of the evaluation setup. The assumption that any brief maps cleanly onto the 17 criteria also sits in the background without much discussion of edge cases.\n\nThe work is aimed at researchers in text-to-3D scene synthesis who want to move beyond object lists. A reader already working on functional constraints or human-centric evaluation would get the most out of it. The central idea is coherent enough on its own terms to merit referee time, even if the current evidence is preliminary.","headline":"Function2Scene reframes scene synthesis around functional briefs and a 17-criteria taxonomy plus iterative geometric-LLM-VLM repair, but the 94.3% preference on 30 cases rests on thin experimental reporting.","tokens_in":2336,"tokens_out":383,"would_cite":false,"duration_ms":16510,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"Function2Scene generates 3D indoor layouts from natural-language descriptions of occupants and their activities rather than from lists of objects.","keywords":["3D indoor scene synthesis","functional specifications","layout generation","text-driven generation","iterative refinement","LLM evaluation","VLM assessment"],"falsifier":"A test set of new design briefs where the generated layouts are not preferred over LLM baselines in human pairwise comparisons, or where the parsing step misses key functional requirements.","tokens_in":2655,"feed_emoji":"🏠","tokens_out":634,"duration_ms":25205,"temperature":0.7,"pith_summary":"Most text-driven 3D scene methods ask what furniture to place. This paper starts instead from functional specifications that describe who will use a room and what they need to do there. The system parses the brief into personas and activities, derives constraints from a taxonomy of 17 criteria across spatial, ergonomic, activity, and environmental dimensions, and then runs an iterative check-and-repair loop that combines geometric measurements, LLM contextual checks, and VLM visual assessments. Experiments on 30 professionally written interior-design cases show the resulting layouts satisfy the stated functional requirements more effectively than recent LLM-based baselines. Human evaluators preferred the outputs in 94.3 percent of pairwise comparisons.","feed_headline":"AI turns occupant activity briefs into 3D room layouts","feed_subtitle":"It outperforms object-list baselines by meeting activity and ergonomic needs in 94 percent of comparisons.","key_machinery":"Tool-augmented iterative check-and-repair loop driven by constraints derived from the 17-criteria functional taxonomy.","core_discovery":"Given a natural-language design brief, Function2Scene parses occupant personas and activities, derives a customized set of functional design constraints from a taxonomy of 17 criteria spanning spatial, ergonomic, activity, and environmental considerations, and uses these constraints to guide layout generation through iterative evaluation and refinement that combines geometric measurements, LLM-based contextual reasoning, and VLM-based visual assessment.","pith_inferences":["The fixed 17-criteria taxonomy could be extended or replaced by learned criteria to handle requirements outside the current set.","Adding physics-based simulation inside the repair loop could validate dynamic movement and reachability beyond static geometric checks.","The same parsing-plus-refinement structure could be applied to outdoor or multi-room functional briefs."],"forward_implications":["Layouts better satisfy the functional requirements of the input briefs compared with object-centric prompts.","Human raters prefer the outputs over recent LLM-based scene synthesis baselines in 94.3 percent of pairwise comparisons.","Scene synthesis is reframed from placing plausible objects to designing spaces that support human use.","The method applies directly to professionally written interior-design cases."],"fun_headline_variants":["Design briefs produce 3D layouts from 17 criteria","Activity-focused briefs guide functional 3D room creation","LLM-VLM loop refines layouts for occupant needs","Text specs yield activity-supporting 3D indoor scenes"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"Natural-language design briefs can be reliably parsed into a fixed taxonomy of 17 criteria that then guide effective layout generation through the iterative check-and-repair process.","fun_headline_variants_meta":{"raw":{"variants":["Design briefs produce 3D layouts from 17 criteria","Activity-focused briefs guide functional 3D room creation","LLM-VLM loop refines layouts for occupant needs","Text specs yield activity-supporting 3D indoor scenes"]},"model":"grok-4.3","cost_usd":0.005737,"raw_usage":{"total_tokens":2740,"prompt_tokens":675,"num_sources_used":0,"completion_tokens":64,"cost_in_usd_ticks":57374500,"prompt_tokens_details":{"text_tokens":675,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":2001,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":675,"tokens_out":64,"duration_ms":14607,"temperature":1.0,"reasoning_tokens":2001,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-28T23:08:57.343079+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A test set of new design briefs where the generated layouts are not preferred over LLM baselines in human pairwise comparisons, or where the parsing step misses key functional requirements.","supporting_citations":[],"review_version":1}