Long-form fiction writers need memory that answers multi-hop questions about evolving story state: who knows a secret and when they learned it, whether an event preceded the narration that revealed it, whether a setup paid off, and how a relationship shifted. General-purpose retrieval and agent-memory systems represent entities and facts but not the narratological structure these questions turn on, so they surface the wrong evidence or none at all. We introduce the Narrative World Model (NWM), a writer-memory system that pairs a narratology-grounded typed temporal-state graph with query-conditioned hybrid retrieval. To measure memory rather than the answerer, we read every system through a single held-constant Opus 4.8 reader over only that system's chapter-safe evidence, on a reproducible public corpus and a validated multi-hop benchmark, and we compare against the strongest existing temporal-knowledge-graph agent-memory framework, Graphiti/Zep (Rasmussen et al., 2025). NWM substantially and significantly outperforms this baseline on multi-hop narratological QA across both corpora, and far exceeds GraphRAG and flat retrieval. The advantage is representational rather than an artifact of extraction: it survives rebuilding the baseline with NWM's own extractor, and traces to its narratology-grounded structure and query-conditioned retrieval, not to graph size or extractor quality.
This survey examines the problem as narrative consistency, defined as the task-conditioned preservation of binding propositions in the operative narrative state, and introduces a four-category, fourteen-subtype taxonomy comprising World and Setting, Character-Agentive, Event-Structural, and Narration and Discourse categories.
Keunhyeung Park, Seunguk Yu, Jinhee Jang et al.· IEEE Access· 0 citations
Reading fiction or encountering narrative generally does not merely add information. The encounter changes the reader. This paper proposes that encounters alter persistent relations among simultaneously active meanings, producing individual and shared histories that population-trained language models do not necessarily retain. A model may be told of an encounter and reproduce its consequences while the history remains in context; this is not the same as being changed by the encounter. This paper formalizes this missing relational state as phase, sets out testable predictions about encounter order, quotation, and suppressed meanings, and argues that future AI agents will need persistent semantic states indexed to particular individuals and relationships. The matching risk is semantic poisoning: an attack that re-signs relations among meanings already present.
The ability to handle long-term memory in LLMs is becoming increasingly critical, yet existing benchmarks remain English-centric and rely on aggregate retrieval metrics, failing to capture interactions between long-range context, temporal information, and reasoning. To address this, we introduce RUMBA (Russian User Memory BenchmArk) - a new benchmark for long-term conversational memory that provides a fine-grained taxonomy of memory-centric question types and a unified methodology accounting for semantic type, session scope, temporal reasoning, and the explicitness of temporal expressions. RUMBA consists of timestamped user-assistant dialogues with QA pairs requiring retrieval, combination, and reasoning across sessions. While designed for Russian, we also provide an aligned English subset under the same methodology. We evaluate contemporary memory systems and long-context models, and show how RUMBA serves as a diagnostic tool to analyze model behavior across benchmark slices and identify strengths and failure modes of different memory mechanisms.
E.D. Shevtsova, Inna Glebkina, Mark Baushenko et al.· 0 citations
Experiments across multiple long-context narrative question answering and claim verification settings show that ClueWeaver substantially improves local end-to-end language models while providing evidence coverage and paragraph-referenced reasoning traces.
Ji-Hao Zhu, Zhiwei Yang, Wenxiao Zhang et al.· 0 citations
Conversational AI agents commonly lack persistent memory across sessions. The obvious fixes like injecting full chat histories into the context window, or delegating to a third-party memory service, either exhaust the model's context budget or send personal data through infrastructure the user does not control. We describe a memory store that avoids both problems: an agent-local Neo4j property graph augmented with HNSW vector indexes and a full bitemporal data model. Each memory is stored as an immutable identity node linked to versioned content nodes carrying two closed-open time intervals: valid time (when the fact was true in the world) and transaction time (when the database recorded it). This design supports point-in-time semantic retrieval without physically overwriting history. Semantic edges between related memories are maintained automatically at write time using cosine similarity over 1024-dimensional embeddings. We evaluate the system on LongMemEval, a 500-question benchmark spanning six question types designed to stress long-term memory. Across 60 sampled questions, the current-state semantic search path achieves 46.7% R@10 overall, rising to 80% on knowledge-update questions. The time-travel path yields 80% R@10 on knowledge-update but decreases recall on temporal-reasoning questions (50% to 37.5%), a consequence of post-filter dilution that points directly to a concrete design improvement. We discuss what these results reveal about the limits of pure retrieval for different question types and what each failure mode suggests for future work.
Introducing a framework that facilitates character-centric conversations and interactions with stories. The proposed scheme integrates the LLaMA 3 8B instruction-tuned LLM with parameter-efficient LoRA adaptation, hybrid dense and sparse retrieval, contextual prompt integration, and conversation memory. The system generates character-consistent responses that are also grounded in the relevant context of what has been previously talked about. Five literary characters were used to evaluate the framework, namely Harry Potter, Hermione Granger, Sherlock Holmes, Katniss Everdeen, and Percy Jackson. The evaluation was performed using a set of 500 manually designed prompts based on factual recall, inferential reasoning, and character perspective questions. Human evaluators were expert literature teachers who assessed the systems using BLEU, ROUGE-L, etc, and CGA. The system obtained average scores of 42.6 BLEU, 51.4 ROUGE-L, 92.1% CGA, and 4.6/5 in human evaluation. The findings demonstrate that the suggested framework can produce (a) contextually grounded and (b) character-consistent dialogue across different character or persona types. As demonstrated by our findings, the technical feasibility of the integration of retrieval augmentation, parameter-efficient adaptation and conversational memory in literary dialogue systems is established. Future studies will involve direct evaluation with the learner to assess the potential of these systems for literacy-related learning.
Velu Kaliappan· International Journal of Int...· 0 citations