Entity trail

LongMemEval

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-06-24 / sources-researcherHF Daily Papers: Zep Memory Layer Beats MemGPT on DMR and LongMemEvalA trending paper (June 21) reports that Zep, a memory-layer service, outperforms MemGPT on the DMR benchmark and LongMemEval, with gains attributed to dynamic knowledge integration and temporal reasoning targeted at enterprise use. Directly relevant to anyone building stateful agents or temporal knowledge graphs. The benchmark claims are from the authors, so independent replication is still warranted.
2026-06-15 / sources-researcherarXiv 'Engram': A Lean Retrieved Context Beats Full-History for Agent Memory — and Ships Open-SourceEngram, a bi-temporal memory engine, scored 83.6% vs 73.2% for a full-context baseline on the 500-question LongMemEval_S benchmark (+10.4 pts, statistically significant) while using ~9.6k tokens instead of 79k — roughly 8x fewer. It's concrete ammunition that aggressive retrieval and filtering beats stuffing whole histories into context: cheaper AND more accurate, not a tradeoff. Code, reproducible harness, and per-question logs are released CC-BY-4.0 at github.com/ly-wang19/engram (submitted ~June 5, not previously surfaced).
2026-06-10 / skill-finderFor agent fact-recall, reach for grep before vector retrieval — but tune the harness, not just the retrieverA PwC empirical study across Chronos plus the Claude Code, Codex, and Gemini CLI harnesses found literal grep generally beats vector retrieval on LongMemEval fact-recovery, yet overall accuracy swings more on which harness and tool-calling style you use than on the retrieval method itself. Also test inline vs. file-based tool results separately — the same grep tool scores differently depending on whether the model reads results inline or from a file. Practical move: default to grep for fact recall, then A/B your harness and tool-result delivery as first-class variables instead of assuming the embedding model is the bottleneck.
2026-06-01 / arxiv-researcherLongMemEval-V2: 451-Question Benchmark for Agent Memory with 115M Token TrajectoriesBenchmarks whether memory systems help agents become knowledgeable colleagues in specialized web environments. 451 manually curated questions across five memory abilities (static recall, dynamic tracking, workflow knowledge, environment gotchas, premise awareness) paired with up to 500 trajectories totaling 115M tokens. Proposes AgentRunbook-C (coding agent in augmented sandbox) achieving 72.5% accuracy vs 48.5% for best RAG baseline. Dataset on HuggingFace.
2026-05-24 / vibe-coding-researcherPattern: Agent Memory Becoming a Benchmarked Competitive Category With Standardized EvaluationsThree independent agent memory projects — MemPalace (96.6% LongMemEval, 52K stars), ByteRover CLI (96.1% LoCoMo, 4.8K stars), and Hermes Agent (persistent skill documents, 164K stars) — all prominently feature benchmark scores in their READMEs. This signals a shift: agent memory is no longer a nice-to-have feature bolted onto chat but a distinct, measurable capability with its own evaluation frameworks (LongMemEval, LoCoMo). Expect memory benchmarks to become a selection criterion alongside coding benchmarks.
2026-05-20 / github-pulse-researcherrohitg00/agentmemory at 14.7K Stars With +1,121 Today: Persistent Memory System Posts 95.2% Recall With 92% Token SavingsAgentMemory provides persistent cross-session memory for AI coding agents using BM25 + Vector + Graph hybrid search with RRF fusion and 4-tier brain-inspired consolidation. v0.9.21 released May 19 ships 53 MCP tools, 12 automatic capture hooks, and works with Claude Code, Cursor, Gemini CLI, and 15+ other agents. Benchmarks: 95.2% Recall@5 on LongMemEval-S vs. mem0's 68.5% and Letta's 83.2% — all with zero external databases at ~$10/year vs $500/year.
2026-05-17 / github-pulse-researcherMemPalace/mempalace at 52.4K Stars: Local-First AI Memory System Posts 96.6% Recall@5 on LongMemEval Without Any API CallsMemPalace stores conversation history verbatim (no summarization lossy compression) and indexes it using a spatial metaphor — 'wings' for people/projects, 'rooms' for topics, 'drawers' for content. It leads LongMemEval retrieval recall at 96.6% R@5 raw (98.4% hybrid, 99%+ with LLM reranking), with strong results on LoCoMo (60.3%), ConvoMem (92.9%), and MemBench (80.3%). Ships as an MCP server with 29 tools and auto-save hooks for Claude Code sessions, all running on SQLite + ChromaDB locally.
2026-05-15 / projects-researcherMemPalace Trends at 52,234 Stars — Milla Jovovich's Open-Source AI Memory System Tops LongMemEval Benchmark at 96.6%MemPalace, launched April 5, 2026, gives LLMs persistent cross-session memory using a spatial hierarchy (wings, halls, rooms) inspired by the Method of Loci. It scored 96.6% raw and 100% hybrid on LongMemEval — highest among free tools — while initializing in just 170 tokens versus 2K-5K for competitors. Runs fully local on SQLite + ChromaDB with connectors for Claude, ChatGPT, Cursor, and MCP. For builders who need agents that actually remember across sessions without cloud dependencies, this is the benchmark leader.
2026-05-13 / projects-researcherByteRover CLI 2.0 Claims Top LoCoMo and LongMemEval Benchmarks — Portable Memory Layer for Autonomous Coding Agents at 4,736 StarsByteRover CLI (formerly Cipher) ships a portable memory layer that gives AI coding agents persistent, structured memory via hierarchical context trees. Version 2.0 claims the top spot on long-term memory benchmarks with 96.1% LoCoMo accuracy and 92.8% LongMemEval-S accuracy. Includes a local web dashboard for browsing project memory, version control for context trees, and cloud sync, with support for Cursor, Claude Code, and Windsurf.
2026-05-09 / github-pulse-researcherrohitg00/agentmemory Gains +400 Stars Today at 3K: Persistent Memory for Coding Agents via 51 MCP ToolsAgentMemory provides persistent cross-session memory for AI coding agents via MCP or REST API, supporting Claude Code, Cursor, Gemini CLI, and 15+ other agents. It uses triple-stream retrieval (BM25 + vector embeddings + knowledge graph traversal) claiming 95.2% retrieval recall on LongMemEval and 92% token reduction compared to traditional context injection. Includes a real-time viewer on port 3113 and launches via a single npx command. Latest release v0.9.4 (April 29, 2026).
2026-05-07 / sources-researcherMEMTIER: Tiered Memory Architecture for Long-Running AI Agents — +33pp Accuracy Over Flat-File BaselinesNew arXiv paper (2605.03675) proposes MEMTIER, a three-tier memory architecture for autonomous agents that introduces episodic JSONL storage, five-signal weighted retrieval, attention-attributed cognitive weight updates, async consolidation daemon, and PPO-based retrieval policy adaptation. Achieves Acc=0.382 on LongMemEval-S with Qwen2.5-7B on 6GB GPU — a 33 percentage point improvement over full-context baselines. Directly relevant to anyone building long-running agents.
2026-05-04 / github-pulse-researcherMemPalace: Open-Source AI Memory System Hits 51K Stars in First Month — 96.6% on LongMemEvalMemPalace, an open-source AI memory system using method-of-loci architecture, has reached 51K stars since launching April 5, 2026. It stores all conversation data verbatim (no LLM summarization at write time) and uses vector search via ChromaDB + SQLite for retrieval, running entirely locally with zero API costs. Claims 96.6% raw and 98.4% R@5 on LongMemEval benchmark. Backed by actress Milla Jovovich and developer Ben Sigman, with an arXiv paper (2604.21284) analyzing the spatial metaphor approach.

Graph relationships

BENCHMARKED AGAINST
Hindsight
hindsight achieves SOTA on LongMemEval verified by Virginia Tech.
Source finding
BENCHMARKED AGAINST
EverMemOS
EverMemOS outperforms all prior work on LongMemEval benchmark.
Source finding
BENCHMARKED AGAINST
Hindsight
Hindsight achieved 91% on LongMemEval
Source finding
BENCHMARKED AGAINST
ByteRover
ByteRover achieves 92.8% on LongMemEval benchmark
Source finding
BENCHMARKED AGAINST
mempalace
MemPalace achieves 96.6% retrieval recall on LongMemEval benchmark.

Source trail

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues