arXiv:2606. 12411v1 Announce Type: cross Abstract: Modern conversational agents condition on an ever-growing dialogue history at each turn, incurring redundant attention and encoding costs that grow with conversation length.
Modern conversational agents condition on an ever-growing dialogue history at each turn, incurring redundant attention and encoding costs that grow with conversation length. Naive truncation or summarization degrades fidelity, while existing context compressors lack cross-turn memory sharing or revision, causing information loss and compounding errors in long dialogues.
arXiv:2601. 07994v5 Announce Type: replace-cross Abstract: Large Language Models (LLMs) increasingly operate over long-form dialogues with frequent topic shifts.
The paper introduces SCALE-QA, a new QA benchmark that tests conversational memory in flat, unsegmented multi‑topic threads by requiring agents to infer which earlier episode supports a later task decision. The dataset contains 3,000 audited questions across ten domains, uses deterministic four‑way multiple‑choice grading, and includes a runtime builder for reproducibility. The authors also propose Temporal‑Semantic Interleaved Memory Reconstruction (TSIM), a hierarchical memory stack that segments turns into coherent episodes and indexes them with deterministic summaries and cluster‑routing views, achieving significant accuracy gains over strong RAG baselines and long‑context LLMs.
arXiv:2608. 03463v1 Announce Type: new Abstract: Long-term memory is essential for LLM-based agents to sustain interactions and reliably leverage distant history.
Large language model agents have shown strong capabilities in generating coherent and contextually appropriate responses, yet robust long-horizon dialogue remains limited by the lack of external memory that is traceable, updatable, and diagnostically transparent. Existing memory-augmented agents often store memories as isolated records or overwritable states, making it difficult to preserve how information originates, evolves, conflicts, or becomes obsolete over time.
CueMem is a cue‑guided framework for long‑term conversational memory that reconstructs query‑relevant dialogue context from compressed memory records. Instead of treating memory units as self‑contained evidence, it extracts fine‑grained cues linked to their source turns and, at query time, expands from these cues over a turn graph to rebuild a compact evidence context. Experiments on LoCoMo and LongMemEval show that CueMem outperforms baseline memory methods, reduces input tokens and latency, and improves long‑term conversational question answering.
arXiv:2606. 03463v1 Announce Type: new Abstract: Conversational AI agents require memory systems that are both scalable and semantically coherent across long interaction horizons.
arXiv:2606. 09659v1 Announce Type: cross Abstract: Long-context language model inference is bottlenecked by memory, as the KV cache grows with context length.
DiaRelay introduces a lightweight adapter that lets large language models maintain a constant‑size dialogue‑level memory for emotion recognition in conversation. It builds on LoRA by adding a Selective Relay Memory Transition that aggregates useful historical evidence into a bounded memory, and a Dual‑axis Relay Memory Read that uses this memory to modulate low‑rank feature transformations. Experiments show DiaRelay achieves state‑of‑the‑art weighted F1 and accuracy on MELD with only 7.1 M additional trainable parameters, while also performing competitively on IEMOCAP.
Long-context language model inference is bottlenecked by memory, as the KV cache grows with context length. Recent techniques to compress the KV cache fall short: they either degrade model quality substantially or require considerable time and compute to compress a single long prompt.
The paper introduces a long‑context recall technique that keeps GPU memory usage nearly constant regardless of context length, without requiring additional training. It reconstructs facts by leveraging residual vectors stored in the LLM’s feed‑forward layers, enabling deterministic retrieval of query‑relevant information without accessing the original document. Experiments demonstrate the method can answer single‑fact questions in two‑million‑token stories, outperforming prior approaches.
arXiv:2608. 09043v1 Announce Type: cross Abstract: Users of modern platforms repeatedly need summaries of recent dialogue, but the window rarely contains enough context to be interpreted on its own.