EgoMemReason is a new benchmark showing that even the best multimodal models achieve only 39.6% accuracy on reasoning tasks that require integrating sparse evidence across days in egocentric video.
hub Canonical reference
Seeing, listening, remembering, and reasoning: A multi- modal agent with long-term memory
Canonical reference. 100% of citing Pith papers cite this work as background.
20
Pith papers citing it
Background
100% of classified citations
hub tools
citation-role summary
background 6
citation-polarity summary
years
2026 20verdicts
UNVERDICTED 20roles
background 6polarities
background 6representative citing papers
DMV-Bench introduces the first interactive benchmark for multimodal-agent visual memory via incidental cue injection on product images, and DualMem, a parallel visual-verbal memory architecture, outperforms baselines across chain lengths 5-50 on two VLMs.
MemLens benchmark shows long-context LVLMs lose accuracy with length while memory agents lose visual fidelity, with multi-session reasoning below 30% for most systems and neither approach solving the task alone.
Omni-Persona benchmark with 18 tasks shows open-source models have audio-visual grounding gaps, RLVR narrows them but leads to conservative outputs, and scale or recall alone fail as diagnostics.
MemCompiler reframes memory use as state-conditioned compilation, delivering relevant guidance via text and latent channels to improve embodied agent performance up to 129% and cut latency 60% versus static injection.
MemCoE learns memory organization guidelines via contrastive feedback and then trains a guideline-aligned RL policy for memory updates, yielding consistent gains on personalization benchmarks.
GRAB-ANNS is a new GPU graph index that achieves up to 240x higher hybrid search throughput via bucket layouts and hybrid intra/inter-bucket edges.
SensorPersona uses LLMs for hierarchical reasoning on longitudinal mobile sensor streams to continually extract stable personas, showing up to 31.4% higher recall and 85.7% win rate over baselines on a 20-user dataset.
MM-Mem distills video input through a hierarchical memory of sensory buffer, episodic stream, and symbolic schema, optimized by a semantic information bottleneck and SIB-GRPO, to achieve SOTA on long-horizon video benchmarks.
TaskMem uses RL in two phases to learn a task-focused memorization policy for multimodal agents, yielding 5.3-7.0% VQA accuracy gains on reformulated streaming benchmarks from VideoMME, EgoLife, and EgoTempo.
FluxMem evolves memory as a heterogeneous graph via three refinement stages and reports consistent state-of-the-art results on LoCoMo, Mind2Web, and GAIA benchmarks.
PVM adds a parallel branch to LVLMs that directly supplies visual embeddings to prevent attention decay over long generated sequences, yielding accuracy gains on reasoning tasks with minimal overhead.
POINTS-Long is a dual-mode multimodal large language model that uses dynamic visual token scaling to retain 97.7-99.7% accuracy on long-form tasks with 1/40 to 1/10th the tokens and supports streaming via detachable KV-cache.
AdecPilot decentralizes administration in edge-cloud multi-agent frameworks by using a UI-agnostic cloud designer and a bimodal edge team with a Hierarchical Implicit Termination protocol, yielding 21.7% higher task success, 37.5% less cloud tokens, and 88.9% lower latency.
FileGram grounds AI agent personalization in file-system behavioral traces via a data simulation engine, a diagnostic benchmark, and a bottom-up memory architecture.
PersonaVLM adds memory extraction, multi-turn retrieval-based reasoning, and personality inference to multimodal LLMs, yielding 22.4% gains on a new long-term personalization benchmark and outperforming GPT-4o.
SpeechLess enables micro-utterance AR interactions by binding prior interactions to personal spatial context for intent extrapolation.
This survey categorizes agentic environments for LLMs by eight attributes and domains, introduces symbolic and neural synthesis paradigms with evaluation, and outlines four agent evolution pathways plus three environment evolution paradigms.
PyraVid is a hierarchical multimodal memory system that structures long videos into pyramids to improve long-horizon reasoning and evidence aggregation.
MACF decouples agent perception budgets from overall video length using latent token collaboration to scale video understanding in MLLMs beyond current limits.
citing papers explorer
-
EgoMemReason: A Memory-Driven Reasoning Benchmark for Long-Horizon Egocentric Video Understanding
EgoMemReason is a new benchmark showing that even the best multimodal models achieve only 39.6% accuracy on reasoning tasks that require integrating sparse evidence across days in egocentric video.
-
DMV-Bench: Diagnosing Long-Horizon Multimodal Agents' Visual Memory with Incidental Cue Injection
DMV-Bench introduces the first interactive benchmark for multimodal-agent visual memory via incidental cue injection on product images, and DualMem, a parallel visual-verbal memory architecture, outperforms baselines across chain lengths 5-50 on two VLMs.
-
MemLens: Benchmarking Multimodal Long-Term Memory in Large Vision-Language Models
MemLens benchmark shows long-context LVLMs lose accuracy with length while memory agents lose visual fidelity, with multi-session reasoning below 30% for most systems and neither approach solving the task alone.
-
Omni-Persona: Systematic Benchmarking and Improving Omnimodal Personalization
Omni-Persona benchmark with 18 tasks shows open-source models have audio-visual grounding gaps, RLVR narrows them but leads to conservative outputs, and scale or recall alone fail as diagnostics.
-
MemCompiler: Compile, Don't Inject -- State-Conditioned Memory for Embodied Agents
MemCompiler reframes memory use as state-conditioned compilation, delivering relevant guidance via text and latent channels to improve embodied agent performance up to 129% and cut latency 60% versus static injection.
-
Learning How and What to Memorize: Cognition-Inspired Two-Stage Optimization for Evolving Memory
MemCoE learns memory organization guidelines via contrastive feedback and then trains a guideline-aligned RL policy for memory updates, yielding consistent gains on personalization benchmarks.
-
GRAB-ANNS: High-Throughput Indexing and Hybrid Search via GPU-Native Bucketing
GRAB-ANNS is a new GPU graph index that achieves up to 240x higher hybrid search throughput via bucket layouts and hybrid intra/inter-bucket edges.
-
SensorPersona: An LLM-Empowered System for Continual Persona Extraction from Longitudinal Mobile Sensor Streams
SensorPersona uses LLMs for hierarchical reasoning on longitudinal mobile sensor streams to continually extract stable personas, showing up to 31.4% higher recall and 85.7% win rate over baselines on a 20-user dataset.
-
From Verbatim to Gist: Distilling Pyramidal Multimodal Memory via Semantic Information Bottleneck for Long-Horizon Video Agents
MM-Mem distills video input through a hierarchical memory of sensory buffer, episodic stream, and symbolic schema, optimized by a semantic information bottleneck and SIB-GRPO, to achieve SOTA on long-horizon video benchmarks.
-
Task-Focused Memorization for Multimodal Agents
TaskMem uses RL in two phases to learn a task-focused memorization policy for multimodal agents, yielding 5.3-7.0% VQA accuracy gains on reformulated streaming benchmarks from VideoMME, EgoLife, and EgoTempo.
-
Rethinking Memory as Continuously Evolving Connectivity
FluxMem evolves memory as a heterogeneous graph via three refinement stages and reports consistent state-of-the-art results on LoCoMo, Mind2Web, and GAIA benchmarks.
-
Persistent Visual Memory: Sustaining Perception for Deep Generation in LVLMs
PVM adds a parallel branch to LVLMs that directly supplies visual embeddings to prevent attention decay over long generated sequences, yielding accuracy gains on reasoning tasks with minimal overhead.
-
POINTS-Long: Adaptive Dual-Mode Visual Reasoning in MLLMs
POINTS-Long is a dual-mode multimodal large language model that uses dynamic visual token scaling to retain 97.7-99.7% accuracy on long-form tasks with 1/40 to 1/10th the tokens and supports streaming via detachable KV-cache.
-
Administrative Decentralization in Edge-Cloud Multi-Agent for Mobile Automation
AdecPilot decentralizes administration in edge-cloud multi-agent frameworks by using a UI-agnostic cloud designer and a bimodal edge team with a Hierarchical Implicit Termination protocol, yielding 21.7% higher task success, 37.5% less cloud tokens, and 88.9% lower latency.
-
FileGram: Grounding Agent Personalization in File-System Behavioral Traces
FileGram grounds AI agent personalization in file-system behavioral traces via a data simulation engine, a diagnostic benchmark, and a bottom-up memory architecture.
-
PersonaVLM: Long-Term Personalized Multimodal LLMs
PersonaVLM adds memory extraction, multi-turn retrieval-based reasoning, and personality inference to multimodal LLMs, yielding 22.4% gains on a new long-term personalization benchmark and outperforming GPT-4o.
-
SpeechLess: Micro-utterance with Personalized Spatial Memory-aware Assistant in Everyday Augmented Reality
SpeechLess enables micro-utterance AR interactions by binding prior interactions to personal spatial context for intent extrapolation.
-
Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application
This survey categorizes agentic environments for LLMs by eight attributes and domains, introduces symbolic and neural synthesis paradigms with evaluation, and outlines four agent evolution pathways plus three environment evolution paradigms.
-
PyraVid: Hierarchical Multimodal Memory for Long-Horizon Video Reasoning
PyraVid is a hierarchical multimodal memory system that structures long videos into pyramids to improve long-horizon reasoning and evidence aggregation.
-
Scaling Video Understanding via Compact Latent Multi-Agent Collaboration
MACF decouples agent perception budgets from overall video length using latent token collaboration to scale video understanding in MLLMs beyond current limits.