Letta retweeted
Interesting work from @Letta_AI. I like the direction of FrontierMemory. Instead of testing whether an agent remembers something, they ask whether the memory actually helps the agent perform better on future tasks.
This article also shows how difficult it is for models to extract reliable, actionable knowledge from past experience.
My guess (not their disclosed architecture) is that Letta is moving from an LLM-based memory swarm toward
===>> a specialized memory learning model, possibly trained with downstream task rewards. Meta RL here can help!
I see this as another approach to building an Agent Companion Model, similar to MemT.
But I think there is a big challenge if they want to make this a commercial product.
As the agent covers more and more tasks, and accumulates more experience, how can they keep inference fast and cheap? If each new task requires more memory retrieval, processing and reasoning, the cost may keep growing with experience.
Moving memory processing offline helps, but doesn’t fully resolve the scaling problem.
I think online plasticity could be the key here.
But how to do this correctly? I don’t think anyone has a convincing solution yet.
This is what I described as the Agent-CM trilemma: Task Breadth × Online Plasticity × Runtime Efficiency.
trueintrinsics.app/writing/a…
Letta retweeted
Earlier this year, i wrote about startups including @Letta_AI and @LangChain exploring new ways for AI models to consolidate memories through something reminiscent of dreaming. Anthropic now says that Claude does something similar wired.com/story/sleeptime-co…
Memory is what allows Claude to learn from experience. In a new Engineering Blog post, we share how “dreaming” helps Claude organize what it’s learned, correct its mistakes, and find what it needs more easily the next time. anthropic.com/research/compo…
Letta retweeted
Very cool blog post from Anthropic that talks about how Claude uses sleep-time compute (from research @Letta_AI) to power learning from experience.
Sleep-time compute is the most powerful form of agent learning today, but many unsolved problems remain!
Memory is what allows Claude to learn from experience. In a new Engineering Blog post, we share how “dreaming” helps Claude organize what it’s learned, correct its mistakes, and find what it needs more easily the next time. anthropic.com/research/compo…
Letta retweeted
sleep-time compute was a big reason I joined @Letta_AI's research team last year. very cool to see it evolve into "dreaming" and become a key primitive for memory!
Memory is what allows Claude to learn from experience. In a new Engineering Blog post, we share how “dreaming” helps Claude organize what it’s learned, correct its mistakes, and find what it needs more easily the next time. anthropic.com/research/compo…
Letta retweeted
Recommended reading and a super-interesting benchmark for measuring continual learning capabilities through memory.
Agent memory is largely unsolved. As agents advance in capabilities like long-horizon work, it's unclear which models or harnesses work best, or which memory systems and types to use.
Opus 5.5 at the top is not a surprise. But it's exciting that open-weight models like GLM-5.3 perform so well on memory generation.
My big question is how this (better memory systems and persistent agent capabilities) translates into more performant, efficient RSI agents and harnesses.
Letta retweeted
Memory is becoming a core part of agent harnesses/products (e.g. Instinct, @Muse, @cognition Devin, @PrimeIntellect Prime Agent, @NousResearch Hermes, @openclaw, @LangChain Deep Agents, @Letta_AI) -- yet despite every agent claiming to "continually learn", there's no benchmark to actually measure that capability.
A big problem is that memory benchmarks today only measure recall (e.g. simple Q/A about past events). They don't measure if/how agents actually learn from experience to get better at work.
At the same time, benchmarks for coding/knowledge work consist of many isolated, unrelated tasks - so they also don't allow for measuring cross-task learning. The context needed for a task is fully specified up-front (very unrealistic).
In practice, both agents and humans work through sequences of many *related* tasks. Experience from past work provides context and learnings that apply to future work. This is why if you're a manager, you have to provide a lot more context when assigning tasks to a brand new employee or external contractor. Meanwhile, a great senior employee might not need any instructions at all.
With FrontierMemory, we adapt existing benchmark tasks for coding and knowledge work so that instead of all the necessary context being provided up-front and within the task, it has to be learned from past experience. This allows us to compare how well different models and harnesses learn over time.
Letta retweeted
FrontierMemory from @Letta_AI is the first benchmark to measure memory based on performance on downstream tasks.
Memory remains one of the biggest unsolved challenges in AI.
FrontierMemory ACTUALLY gives us a way to actually measure and compare memory capabilities across different harnesses and models.
Letta retweeted
Truly capable agents must learn from experience, not merely recall facts. The FrontierMemory benchmark evaluates agent memory directly by measuring its impact on downstream task performance.
Letta retweeted
Hey, if you're still using @Letta_AI Classic like I am, you need an ADE (as Letta has EOL'd and retired the online version).
I'm maintaining the Sanctum branch of both Letta and the ADE here:
github.com/sanctumos/letta
github.com/sanctumos/ade
Letta retweeted
Using Muse to dispatch work to my @Letta_AI cloud agent, which can be orchestrated via MCP
MCP is the real A2A
Letta retweeted
It's AI Infra night at @Letta_AI HQ!
Here's Letta cofounder @sarahwooders showing off Letta, cloud sandboxes, Slack, and teleportation from her laptop to the cloud.
Letta retweeted
Mods in @Letta_AI are very cool, and they work on cloud agents too! (you just need to move the mod into the agent's memory so it becomes part of the agent-owned context)
letta.com/blog/introducing-m…
Letta retweeted
I have learned a lot about agent design and more through these office hours sessions @just_cameron hosts weekly for @Letta_AI
LETTA OFFICE HOURS TODAY AT 11:30 AM PACIFIC
Come hang out on our Discord and learn about recent developments in @Letta_AI: cloud agents, the Agent SDK, shared memory, OpenRouter OAuth, and MORE.
Bring questions, comments, compliments, and high-fives.
Link to join below 👇
ALT Text on a dark background announces office hours with bold letters, featuring a small red square in the corner.
Letta retweeted
LETTA OFFICE HOURS TODAY AT 11:30 AM PACIFIC
Come hang out on our Discord and learn about recent developments in @Letta_AI: cloud agents, the Agent SDK, shared memory, OpenRouter OAuth, and MORE.
Bring questions, comments, compliments, and high-fives.
Link to join below 👇
ALT Text on a dark background announces office hours with bold letters, featuring a small red square in the corner.
Letta retweeted
this is cool because it lets you collect your own coding traces (literal gold) via multiple harnesses
Letta retweeted
Our already very smart support agent just got smarter (running on GPT 5.6 sol)
I also love how you can ask @Letta_AI agents to reconfigure themselves - like reducing the max context size.
Letta retweeted
We renamed Letta Code to Letta Agent.
It's general purpose. Coworkers, colleagues, companions, whatever.
Build agents that actually learn.