{"data":{"event":{"id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","slug":"what-do-current-systematic-generalization-tasks-miss-a-reasoning-centere-2bbd9e9eca","title":"What Do Current Systematic Generalization Tasks Miss? A Reasoning-Centered Analysis","short_summary":"arXiv:2609.19212v1 Announce Type: new \nAbstract: Systematic generalization, the ability to solve novel problems by recombining known atomic elements, is central to human intelligence but difficult to study rigorously under controlled settings. Existing studies therefore rely on simplifications such as approximately linear action composition, productivity-based tests, and action-explicit goals, which make systematic generalization easier to study but omit some essential aspects of this capability. To characterize what these simplifications miss, we adopt a reasoning-centered lens and introduce ","full_description":"arXiv:2609.19212v1 Announce Type: new \nAbstract: Systematic generalization, the ability to solve novel problems by recombining known atomic elements, is central to human intelligence but difficult to study rigorously under controlled settings. Existing studies therefore rely on simplifications such as approximately linear action composition, productivity-based tests, and action-explicit goals, which make systematic generalization easier to study but omit some essential aspects of this capability. To characterize what these simplifications miss, we adopt a reasoning-centered lens and introduce TranSGrid, a testbed that brings deductive, inductive, and abductive reasoning together within a unified task. Experiments with seven Transformers on 4,800 TranSGrid instances show that all models perform much worse on TranSGrid than on a held-out test set: the largest model solves 79.6% of the test set, but only 55.3% of TranSGrid and 15.8% of the hardest subset. The gap remains within the training length range, showing that productivity alone is not sufficient to evaluate systematic generalization. Additionally, we reintroduce the other two simplifications into TranSGrid: one variant makes actions compose almost linearly (reducing the inductive demand), the other makes goals action-explicit (reducing the abductive one). In both, solve rates return to roughly the test set level, showing that either simplification alone is enough to reduce TranSGrid to an ordinary held-out test set. Together, our results show that existing tasks reduce either or both of the inductive and abductive demands, and that comprehensively measuring systematic generalization requires a task that involves all three forms of reasoning.\n\nUpdate (2026-09-18): Characterizing Web Search by Conversational LLM Agents: From Search Decisions and Strategies to Results and Responses. arXiv:2609.19244v1 Announce Type: new \nAbstract: Conversational LLM agents increasingly rely on Web search, yet the end-to-end lifecycle of agentic search remains poorly understood. We present the first study of Web search across four major conversational platforms (ChatGPT, Claude, Grok, and DeepSeek), combining real-world user interactions (invivo) with controlled experiments using the same platform's models by their APIs (invitro). We investigate the quality of agentic decisions to invoke Web search, their strategies to formulate queries, the potential domain preferences in the search results they receive, and the choices they make when transforming search results into grounded responses. We find that Web-search decisions vary substantially across platforms and models, while more frequent Web-search invocation does not necessarily yield better response quality. We further show that conversational agents employ different complex querying strategies and that platform specific search engines return search results from their preferred domains.\n\nUpdate (2026-09-18): MAGS: Multi-agent Auto-formalization Guarantees Safety for Agentic Outputs. arXiv:2609.19391v1 Announce Type: new \nAbstract: LLM coding agents now generate complex programs at a scale that makes thorough human review increasingly difficult, raising the risk of safety and security failures. Common approaches, including fuzz testing, static analysis, and LLM-as-a-Verifier, can detect many failures but struggle to cover all possible edge cases. Formal verification addresses this by providing machine-checkable guarantees over specified properties, but traditionally demands substantial manual specification and proof engineering. We introduce a unified multi-agent framework, MAGS, that generates executable programs with formal safety guarantees, using Dafny as a verification-aware intermediate representation where safety properties can be mechanically checked. MAGS formalizes and freezes human-audited APIs and safety requirements, translates generated code into Dafny, repairs violations using verifier feedback, and compiles verified programs back into executable code. We evaluate MAGS on 100 CUDA kernels, 100 terminal scripts, and 20 robotic-arm tasks. Across all 2\n\nUpdate (2026-09-18): Compositional Reasoning in Language Models under Reinforcement Learning Post-Training. arXiv:2609.19465v1 Announce Type: new \nAbstract: Compositional reasoning is critical for real-world problem solving: since training data is necessarily limited, models must generalize by composing learned skills in new ways. While post-training methods such as reinforcement learning (RL) have substantially improved the reasoning abilities of language models (LMs), their effects on compositional reasoning remain less well understood. We propose a dependency-graph framework to formalize compositional reasoning, yielding three levels of compositionality with increasing complexity. Empirically, we instantiate this framework with data-structure tasks, which provide deterministic reward computation and clear compositional structure. We find a consistent decomposed-to-composed asymmetry: decomposed-skill training does not reliably transfer to composed tasks, whereas composed-task training transfers more readily back to decomposed tasks. We provide theoretical explanation for this asymmetry, and further evaluate compositional generalization under length extrapolation, structural di\n\nUpdate (2026-09-18): An Architecture for Long-Horizon Agents: Levels, Ticks and Cascaded Intelligence. arXiv:2609.19519v1 Announce Type: new \nAbstract: Language-model agents are increasingly asked to carry out work spanning days or weeks, such as an operations remediation or a research programme. Such a task outlives any context window, any process and any interval at which a person can attend. In this paper, we argue that a long-horizon agent must run continually without forgetting before it can learn continually. This ability lies in the harness around the model rather than in the model itself. We derive seven bottlenecks from the long-horizon setting and answer them with a hierarchical architecture of three parts: (i) levels indexed by time scale, each keeping a bounded file summarising the level below; (ii) a clocked tick as the unit of autonomous action; and (iii) cascaded intelligence, where work is escalated to a more capable model only after failing review. We report on a ten-day campaign in which an agent built on this architecture reproduced a published reinforcement-learning result with a human attending once a day, and show (1) the agent kept the thread across every c\n\nUpdate (2026-09-18): EconSkills: Studying Skill Transfer and Retrieval for Web Agents on Live Economic Data. arXiv:2609.19523v1 Announce Type: new \nAbstract: Web agents often revisit the same sites, yet most evaluations discard the procedures learned in earlier successful interactions. We introduce EconSkills, a skill library and evaluation framework that distills verified EconWebArena trajectories into parameterized standard operating procedures for retrieving live economic data. Each skill records its scope, navigation procedure, site-specific guidance, verification checks, and recovery steps while replacing source-instance values with placeholders. EconSkills separates two questions: whether a known relevant procedure transfers to a held-out task, and whether an agent can retain that benefit when selecting from a library. In controlled transfer, matched skills improve success over no-skill prompting and require fewer steps on paired successes, while abstraction is substantially more effective than replaying raw trajectories. At library scale, retrieval is competitive with the no-skill baseline overall and performs best on directly covered tasks; coverage-stratified outcomes sh\n\nUpdate (2026-09-18): When Hiring Becomes Agent-Mediated: Evaluating Access and Recurrence in Two-Agent R\\'esum\\'e Screening. arXiv:2609.19530v1 Announce Type: new \nAbstract: Hiring is bilateral: employers assess fit, while candidates present and defend evidence of their qualifications. Yet r\\'esum\\'e screening, the first gate, is commonly automated as a static, one-call judgment over a r\\'esum\\'e-job pair. We study a two-agent alternative in which employer-side and candidate-side agents represent these roles, exchange evidence, and update their judgments before deciding who advances. We compare procedures on 600 constructed r\\'esum\\'e-job pairs using GPT-5.5 and Claude Opus 4.7. Two-agent screening advances more applications (33.3% to 39.3% for GPT-5.5; 34.0% to 35.5% for Opus 4.7). Across three runs on the common 191-pair borderline pool, pass-instance rates rise from 4.5% to 26.2% and from 6.5% to 16.1%, respectively. This is not a uniform relaxation: two-agent screening rejects applications one-call advances, changing decisions in both directions. At similar pass volumes, the procedures advance different applications, and no one-call threshold recovers applications consistentl\n\nUpdate (2026-09-18): SIMLIFE: Pattern Understanding for Long-Horizon Human-Agent Partnership. arXiv:2609.19610v1 Announce Type: new \nAbstract: Understanding humans over long horizons requires agents to infer not only what people need in the moment, but also how routines form, why they repeat, and when they change. We introduce SimLife, a scalable platform for simulating long-term household life with rich visual observations, ground-truth action logs, and synthetic dialogues with audio. Built on SimLife, SimLife-BP evaluates long-context pattern understanding: the ability to infer latent behavioral rules from weeks or months of everyday observations. The benchmark contains 106 episodes averaging 15.49 hours and 38.57 in-game days, and 1,439 question-answer pairs. Each task probes direct, counterfactual, noisy, and inverse reasoning under different levels of rule hints. Evaluating frontier models and architectures, we find that current models often achieve surface-level prediction without comprehensive rule understanding, rely on frequency-based heuristics rather than if-then reasoning over evidence, and struggle to adapt when behavioral patterns change. These findings suggest that\n\nUpdate (2026-09-18): ScientistTwo: Pioneering the Human Knowledge Frontier with Autonomous AI. arXiv:2609.19644v1 Announce Type: new \nAbstract: Scientific discovery is defined by the ability to identify the boundaries of existing knowledge and venture into unexplored territory. The ultimate vision for AI in science is problem-driven autonomous research: given a fundamental challenge by a human expert, the AI independently navigates the scientific landscape, uncovers theoretical and empirical bottlenecks, and systematically expands the frontier of knowledge. In this paper, we introduce ScientistTwo, a fully autonomous multi-agent framework designed to realize this vision. Specifically, ScientistTwo takes an initial problem as input, establishes state-of-the-art baselines, formulates novel hypotheses, and coordinates specialized agents to orchestrate an end-to-end discovery cycle without human intervention. Moreover, the framework rigorously conducts experiments using diverse datasets and metrics, refines methodologies through automated ablation studies, and validates research findings via a closed-loop simulated peer-review rebuttal engine. To evaluate ScientistTwo's capabilities \n\nUpdate (2026-09-22): A Survey on the Linear Representation Hypothesis. arXiv:2609.22695v1 Announce Type: new \nAbstract: The term \"linear representation hypothesis\" (LRH) has appeared across diverse subfields of artificial intelligence, neuroscience, and cognitive science. But previous works have not consistently treated the LRH as a falsifiable scientific hypothesis; we analyze these inconsistencies and examine their implications for how prior theoretical and methodological results should be interpreted. Based on this analysis, we argue that claims regarding linear representations become well-defined only through careful examination of the model, representation location, feature definition, and evaluation dataset. We therefore propose a more rigorous formalization of the LRH that makes these dependencies explicit and allows the hypothesis to be evaluated as a falsifiable scientific claim. Finally, we identify some non-trivial open problems that warrant further attention from the research community.\n\nUpdate (2026-09-23): 4DGS-JEPA: Temporally Compositional Joint-Embedding Prediction for Dynamic Gaussian Splatting. arXiv:2609.25036v1 Announce Type: new \nAbstract: Dynamic Gaussian Splatting provides an explicit representation of evolving 3D scenes, but existing approaches are primarily optimized for reconstruction, future-state generation, or rendering rather than for learning reusable predictive dynamics. We propose 4DGS-JEPA, a Gaussian-native joint-embedding predictive architecture for causal multi-horizon prediction over dynamic Gaussian scenes. The model uses a hierarchical scene-, motion-group-, and Gaussian-level representation together with a horizon-conditioned transition operator that supports both direct prediction and recursive rollout. Its central principle is temporal composition: different chronological transition paths reaching the same future endpoint should produce compatible predictive states. Endpoint and multi-horizon path supervision anchor these predictions to future target embeddings, while a selective geometry decoder and geometry-level composition ground the learned dynamics in consistent group motion and Gaussian geometry without requiring complete fu","ledger_type":"benefit","primary_domain_id":"8f1af1b9-7302-4b51-aae9-fdfac04d158a","event_status":"provisional","event_date":"2026-09-18T00:00:00.000Z","discovery_date":"2026-09-18T00:00:00.000Z","first_published_date":"2026-09-18T00:00:00.000Z","last_reviewed_date":"2026-09-23T00:00:00.000Z","geographic_scope":"International","affected_population":null,"base_impact_tier":1,"base_score":"1.00","attribution_multiplier":"0.7500","evidence_multiplier":"0.5000","realization_multiplier":"0.4000","durability_multiplier":"0.5000","current_event_score":"0.075000","confidence_level":"low","score_explanation":"Ingest update: independent source raised evidence; added a follow-up source. Score remains provisional until review. Follow-up ingest: independent source raised evidence; added a follow-up source. Follow-up ingest: independent source raised evidence; added a follow-up source. Follow-up ingest: independent source raised evidence; added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: raised attribution; raised realization; named Claude Opus 4.7; added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: added a follow-up source. Follow-up ingest: added a follow-up source.","methodology_version_id":"8745541b-0f8e-4d5a-9b47-5291d3813a8d","original_methodology_version_id":"8745541b-0f8e-4d5a-9b47-5291d3813a8d","published_at":"2026-09-18T04:00:53.418Z","created_at":"2026-09-18T04:00:53.418Z","updated_at":"2026-09-23T04:01:03.527Z","flags":[],"domain_name":"Biology","domain_slug":"biology","methodology_version":"0.2"},"contributions":[{"id":"be85000b-62e4-436c-86c0-58585578551b","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":"5bf21de4-7735-400c-b5c2-67ec836f5c53","role_description":"Remapped from unspecified after ingest named Claude Opus 4.7.","attribution_multiplier":"0.7500","credit_share":"1.0000","contribution_score":"0.075000","attribution_rationale":"Follow-up source named Claude Opus 4.7.","attribution_confidence":"medium","first_used_date":null,"model_version_if_known":null,"review_status":"approved","created_at":"2026-09-18T04:00:57.869Z","model_slug":"claude-opus-4-7","model_name":"Claude Opus 4.7","identity_class":"named_release","is_internal":false,"family_name":"Claude","family_slug":"anthropic-claude","organization_name":"Anthropic","organization_slug":"anthropic"}],"sources":[{"id":"03460cec-10c1-43ab-ad38-7edae9067c79","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.19212","canonical_url":"https://arxiv.org/abs/2609.19212","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T04:00:53.433Z","title":"What Do Current Systematic Generalization Tasks Miss? A Reasoning-Centered Analysis","excerpt":"arXiv:2609.19212v1 Announce Type: new \nAbstract: Systematic generalization, the ability to solve novel problems by recombining known atomic elements, is central to human intelligence but difficult to study rigorously under controlled settings. Existing studies therefore rely on simplifications such as approximately linear action composition, productivity-based tests, and action-explicit goals, which make systematic generalization easier to study but omit some essential aspects of this capability","content_hash":null,"source_reliability_class":"medium","is_primary_source":true,"is_independent":true,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T04:00:53.433Z"},{"id":"ee8f7458-b36d-444d-8c35-156dc6ac4691","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.19244","canonical_url":"https://arxiv.org/abs/2609.19244","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T04:00:53.692Z","title":"Characterizing Web Search by Conversational LLM Agents: From Search Decisions and Strategies to Results and Responses","excerpt":"arXiv:2609.19244v1 Announce Type: new \nAbstract: Conversational LLM agents increasingly rely on Web search, yet the end-to-end lifecycle of agentic search remains poorly understood. We present the first study of Web search across four major conversational platforms (ChatGPT, Claude, Grok, and DeepSeek), combining real-world user interactions (invivo) with controlled experiments using the same platform's models by their APIs (invitro). We investigate the quality of agentic decisions to invoke Web","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T04:00:53.692Z"},{"id":"76a8b3cc-e474-4350-a5c1-fa12d8a8d215","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.19391","canonical_url":"https://arxiv.org/abs/2609.19391","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T04:00:54.181Z","title":"MAGS: Multi-agent Auto-formalization Guarantees Safety for Agentic Outputs","excerpt":"arXiv:2609.19391v1 Announce Type: new \nAbstract: LLM coding agents now generate complex programs at a scale that makes thorough human review increasingly difficult, raising the risk of safety and security failures. Common approaches, including fuzz testing, static analysis, and LLM-as-a-Verifier, can detect many failures but struggle to cover all possible edge cases. Formal verification addresses this by providing machine-checkable guarantees over specified properties, but traditionally demands ","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T04:00:54.181Z"},{"id":"ee7f3e6a-be27-40e3-8b88-a603cfb3db61","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.19465","canonical_url":"https://arxiv.org/abs/2609.19465","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T04:00:55.140Z","title":"Compositional Reasoning in Language Models under Reinforcement Learning Post-Training","excerpt":"arXiv:2609.19465v1 Announce Type: new \nAbstract: Compositional reasoning is critical for real-world problem solving: since training data is necessarily limited, models must generalize by composing learned skills in new ways. While post-training methods such as reinforcement learning (RL) have substantially improved the reasoning abilities of language models (LMs), their effects on compositional reasoning remain less well understood. We propose a dependency-graph framework to formalize compositio","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T04:00:55.140Z"},{"id":"340940c7-697c-4b94-ac8b-65372cbf2589","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.19519","canonical_url":"https://arxiv.org/abs/2609.19519","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T04:00:56.739Z","title":"An Architecture for Long-Horizon Agents: Levels, Ticks and Cascaded Intelligence","excerpt":"arXiv:2609.19519v1 Announce Type: new \nAbstract: Language-model agents are increasingly asked to carry out work spanning days or weeks, such as an operations remediation or a research programme. Such a task outlives any context window, any process and any interval at which a person can attend. In this paper, we argue that a long-horizon agent must run continually without forgetting before it can learn continually. This ability lies in the harness around the model rather than in the model itself.","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T04:00:56.739Z"},{"id":"c4a18b0d-8d74-44e7-a700-9bca215e4acd","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.19523","canonical_url":"https://arxiv.org/abs/2609.19523","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T04:00:56.991Z","title":"EconSkills: Studying Skill Transfer and Retrieval for Web Agents on Live Economic Data","excerpt":"arXiv:2609.19523v1 Announce Type: new \nAbstract: Web agents often revisit the same sites, yet most evaluations discard the procedures learned in earlier successful interactions. We introduce EconSkills, a skill library and evaluation framework that distills verified EconWebArena trajectories into parameterized standard operating procedures for retrieving live economic data. Each skill records its scope, navigation procedure, site-specific guidance, verification checks, and recovery steps while r","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T04:00:56.991Z"},{"id":"fb029750-9f1a-4ac0-8d6b-faec9af715e9","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.19530","canonical_url":"https://arxiv.org/abs/2609.19530","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T04:00:57.856Z","title":"When Hiring Becomes Agent-Mediated: Evaluating Access and Recurrence in Two-Agent R\\'esum\\'e Screening","excerpt":"arXiv:2609.19530v1 Announce Type: new \nAbstract: Hiring is bilateral: employers assess fit, while candidates present and defend evidence of their qualifications. Yet r\\'esum\\'e screening, the first gate, is commonly automated as a static, one-call judgment over a r\\'esum\\'e-job pair. We study a two-agent alternative in which employer-side and candidate-side agents represent these roles, exchange evidence, and update their judgments before deciding who advances. We compare procedures on 600 const","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T04:00:57.856Z"},{"id":"2cca5c0d-e53b-491a-88ef-f55cf8b110e4","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.19610","canonical_url":"https://arxiv.org/abs/2609.19610","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T04:00:58.478Z","title":"SIMLIFE: Pattern Understanding for Long-Horizon Human-Agent Partnership","excerpt":"arXiv:2609.19610v1 Announce Type: new \nAbstract: Understanding humans over long horizons requires agents to infer not only what people need in the moment, but also how routines form, why they repeat, and when they change. We introduce SimLife, a scalable platform for simulating long-term household life with rich visual observations, ground-truth action logs, and synthetic dialogues with audio. Built on SimLife, SimLife-BP evaluates long-context pattern understanding: the ability to infer latent ","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T04:00:58.478Z"},{"id":"90370d32-4cbf-4b01-9fa7-1f67cc572cdd","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.19644","canonical_url":"https://arxiv.org/abs/2609.19644","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-18T00:00:00.000Z","retrieved_at":"2026-09-18T04:00:59.307Z","title":"ScientistTwo: Pioneering the Human Knowledge Frontier with Autonomous AI","excerpt":"arXiv:2609.19644v1 Announce Type: new \nAbstract: Scientific discovery is defined by the ability to identify the boundaries of existing knowledge and venture into unexplored territory. The ultimate vision for AI in science is problem-driven autonomous research: given a fundamental challenge by a human expert, the AI independently navigates the scientific landscape, uncovers theoretical and empirical bottlenecks, and systematically expands the frontier of knowledge. In this paper, we introduce Sci","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-18T04:00:59.307Z"},{"id":"12bb554d-0374-4aee-bfcb-3fb82a2ac3ce","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.22695","canonical_url":"https://arxiv.org/abs/2609.22695","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-22T00:00:00.000Z","retrieved_at":"2026-09-23T03:01:08.718Z","title":"A Survey on the Linear Representation Hypothesis","excerpt":"arXiv:2609.22695v1 Announce Type: new \nAbstract: The term \"linear representation hypothesis\" (LRH) has appeared across diverse subfields of artificial intelligence, neuroscience, and cognitive science. But previous works have not consistently treated the LRH as a falsifiable scientific hypothesis; we analyze these inconsistencies and examine their implications for how prior theoretical and methodological results should be interpreted. Based on this analysis, we argue that claims regarding linear","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-23T03:01:08.718Z"},{"id":"2e17d07d-1a39-4ea9-859d-4b92d5bc3316","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","url":"https://arxiv.org/abs/2609.25036","canonical_url":"https://arxiv.org/abs/2609.25036","source_type":"preprint","publisher":"arXiv cs.AI","author":null,"publication_date":"2026-09-23T00:00:00.000Z","retrieved_at":"2026-09-23T04:01:03.517Z","title":"4DGS-JEPA: Temporally Compositional Joint-Embedding Prediction for Dynamic Gaussian Splatting","excerpt":"arXiv:2609.25036v1 Announce Type: new \nAbstract: Dynamic Gaussian Splatting provides an explicit representation of evolving 3D scenes, but existing approaches are primarily optimized for reconstruction, future-state generation, or rendering rather than for learning reusable predictive dynamics. We propose 4DGS-JEPA, a Gaussian-native joint-embedding predictive architecture for causal multi-horizon prediction over dynamic Gaussian scenes. The model uses a hierarchical scene-, motion-group-, and G","content_hash":null,"source_reliability_class":"medium","is_primary_source":false,"is_independent":false,"is_peer_reviewed":false,"archived_url":null,"created_at":"2026-09-23T04:01:03.517Z"}],"claims":[{"id":"835e0ae2-90cf-413f-afb3-3c02b902ec0c","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","claim_text":"What Do Current Systematic Generalization Tasks Miss? A Reasoning-Centered Analysis","claim_type":"outcome","claim_status":"supported","confidence_score":null,"created_at":"2026-09-18T04:00:53.440Z","updated_at":"2026-09-18T04:00:53.440Z"}],"revisions":[{"id":"ebf74e22-2db8-4594-8277-667989a9982b","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.000000","new_score":"0.001000","previous_factors":{},"new_factors":{"tier":1,"evidence":0.1,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Auto-published from news ingest.","trigger_type":"news_ingest","trigger_source_ids":null,"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T04:00:53.447Z"},{"id":"303200b5-f366-4080-abbf-55e3f332c2bb","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.001000","new_score":"0.002000","previous_factors":{"tier":1,"evidence":0.1,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.2,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["ee8f7458-b36d-444d-8c35-156dc6ac4691"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T04:00:53.713Z"},{"id":"3a70beb5-9fbf-4aa2-987d-cde729ab7e21","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.002000","new_score":"0.003000","previous_factors":{"tier":1,"evidence":0.2,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.3,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["76a8b3cc-e474-4350-a5c1-fa12d8a8d215"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T04:00:54.202Z"},{"id":"54daff2a-96eb-4ec1-8bf2-6d5196fcbd33","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.003000","new_score":"0.004000","previous_factors":{"tier":1,"evidence":0.3,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.4,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["ee7f3e6a-be27-40e3-8b88-a603cfb3db61"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T04:00:55.165Z"},{"id":"0c741d43-4ca2-4162-a952-acd23b0fa948","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.004000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.4,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: independent source raised evidence; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["340940c7-697c-4b94-ac8b-65372cbf2589"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T04:00:56.762Z"},{"id":"db35a4e1-0c7a-4882-a4be-6f557b283842","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.005000","new_score":"0.005000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["535014f7-c941-4d39-aa9b-8f9c1d3d07b7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["c4a18b0d-8d74-44e7-a700-9bca215e4acd"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T04:00:57.005Z"},{"id":"3352a2ba-f10c-48c4-bb3f-09e2e63ace48","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.005000","new_score":"0.075000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.1,"realization":0.2},"new_factors":{"tier":1,"models":["Claude Opus 4.7"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.75,"realization":0.4},"change_reason":"Ingest update: raised attribution; raised realization; named Claude Opus 4.7; added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["fb029750-9f1a-4ac0-8d6b-faec9af715e9"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T04:00:57.892Z"},{"id":"8cea51cc-9676-4e04-a977-5770fe84a0f1","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.075000","new_score":"0.075000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.75,"realization":0.4},"new_factors":{"tier":1,"models":["5bf21de4-7735-400c-b5c2-67ec836f5c53"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.75,"realization":0.4},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["2cca5c0d-e53b-491a-88ef-f55cf8b110e4"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T04:00:58.494Z"},{"id":"cd229290-74b0-49e2-8fbb-97e2794a1c82","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.075000","new_score":"0.075000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.75,"realization":0.4},"new_factors":{"tier":1,"models":["5bf21de4-7735-400c-b5c2-67ec836f5c53"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.75,"realization":0.4},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["90370d32-4cbf-4b01-9fa7-1f67cc572cdd"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-18T04:00:59.339Z"},{"id":"38b35176-1c89-40bc-9992-e84bc3b35916","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.075000","new_score":"0.075000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.75,"realization":0.4},"new_factors":{"tier":1,"models":["5bf21de4-7735-400c-b5c2-67ec836f5c53"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.75,"realization":0.4},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["12bb554d-0374-4aee-bfcb-3fb82a2ac3ce"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-23T03:01:08.738Z"},{"id":"30e68d44-3b5b-4e41-bc3c-3755f79627ce","event_id":"3551fc50-9bf6-4bdd-821e-b3c65e1f0879","model_id":null,"previous_score":"0.075000","new_score":"0.075000","previous_factors":{"tier":1,"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.75,"realization":0.4},"new_factors":{"tier":1,"models":["5bf21de4-7735-400c-b5c2-67ec836f5c53"],"evidence":0.5,"base_score":1,"durability":0.5,"attribution":0.75,"realization":0.4},"change_reason":"Ingest update: added a follow-up source.","trigger_type":"ingest_update","trigger_source_ids":["2e17d07d-1a39-4ea9-859d-4b92d5bc3316"],"reviewer_id":null,"review_status":"published","created_at":"2026-09-23T04:01:03.534Z"}],"secondary":[]},"methodology_version":"0.1"}