Join the discussion
Write your take first — we'll ask for email only when you're ready to publish.
- Hacker News
- I thought in this day and age "world model" also includes robo arm training data and robot arm benchmarks
- Never heard that.
A world model builds itself a model of the world in which it can simulate an outcome.
In best case its not depending on robotic, otherwise it will be quite limiting for what you can use it.
You can imagine what happens when you write your boss a very inappropriate email, you don't need robotic arms for it.
by blenklo - The labels of the very first chart (figure 1, bottom left) are obviously wrong which casts a doubt on the entire paper.by Tepix
- This label?
> Figure 1: Overview of Qwen-AgentWorld. Top: Qwen-AgentWorld is a unified native language world model across seven domains. Bottom: We explore two complementary strategies for applying world modeling to enhance language agents (mainly using the 35B-A3B model as agent): Decouple and Unify , where the world model serves as the environment simulator and agent foundation model, respectively.
Where is the mistake?
by dudisubekti - The benchmarks here are confusing at best. Am I reading correctly that this model is essentially as good or better than all frontier models right now?by aliljet
- Benchmarks in general are a little iffy, the whole industry is going off of vibes anyways. Can't decide before trying it outby blourvim
- I believe the benchmark listed is about simulating the environment for the various tasks, rather than doing them. It seems that the point of this model is to generate sim data to improve other models withby anana_
- by trilogic
- This questions the nature of banning SOTA models like fable deeply.
As simpler models with better simulated context will be able to more practically execute than SOTAs without such training.
To me this says we should open fable up for defensive reasons rather than fear offensive use. SOTA models will be continuously outmatched by better technique lower grade models with better context techniques like this plus longer walks and deeper inference.
Now you might says SOTAs then could use that and go even further… but how are you going to keep that cat in the bag anyways?
by tsunamifury - I understand what the model is doing. I am struggling to understand where this is going to fit in a workflow. I understand a big gap is that any LLM based ai agent isn't aware of the consequences of its actions because it barely understands the future state its actions will have, hence this model that can.
So, is this like a bolt on where you have an agent powered by an LLM, then the world model reviews the action it wants to take, and the agent confirms this is the intention? Like is this to augment an existing agent with additional capabilities?
by androiddrew - It looks like the purpose of this model is to i. generate environmental sim data for doing RL on other models or ii. act as a foundation model (they trained it to select actions as well as predicting the next state in the same loop?)
Either way, neither are intended for end consumers.
by anana_ - > I understand a big gap is that any LLM based ai agent isn't aware of the consequences of its actions because it barely understands the future state its actions will have, hence this model that can.
These are probably equivalent. Ie, awareness of consequences is the same as understanding the future state. And the present state for that matter, I don't see how someone could be said to understand something if they can't predict the consequences of interacting with it. It is forcing the model to develop a more complex internal world model.
by roenxi - I think the next movement is heading to multi model orchestration.
https://developer.nvidia.com/blog/train-small-orchestration-...
by pulkas - Yeah, this model actually seems like it could handle the job now with good prompting.
- Eli5? What is this compared to a regular llm assistant model like the base qwen?by psc007
- Same thing, but qwen has decided to rebrand certain LLMs that were trained slightly differently as "world models". Despite the fact that "world model" typically means !LLM.by Freedumbs
- A regular LLM acts as a "policy," mapping a current state to a specific action (states → actions). Their new LLM acts as a "world model," mapping a current state and a chosen action to a predicted future state ((states, actions) → subsequent states). Instead of deciding "what to do," its explicit objective is to predict the exact environment observation that will result from the interaction history and the agent's current action.
I assumed at first that it was trained on synthetic data, but they actually went and deployed real physical hosts and virtual machines (e.g. Ubuntu, macOS, and Android) and browsers. They ran agentic systems on these continuously and recorded the actual, real-world interactions.
So it's an LLM that infers next state, or outcome,as structured data e.g. literal HTML code, UI view hierarchies, or accessibility trees.
by gavmor - The smaller of the two models is open weights and available on Huggingface:by adrian_b
- Give it a day or two and the 'unsloth' people will probably publish a Q6 and Q8 (maybe Q8XL?) quantization in GGUF format for llama-server and other users.by walrus01
- I tried to run it but seems like it is either broken or it does not work on dockerized llama.cpp:
0.01.865.326 E llama_model_load: error loading model: missing tensor 'blk.40.attn_norm.weight'
by npodbielski - I'm a fan of this direction. For me the most interesting use case for these world models isn't even training, it's verification. If this thing or some idealized version of it can actually reliably simulate state transitions, could you use it to verify an agent's execution path against hard constraints and replace/eclipse LLMs-as-a-judge?by dippogriff
- Well if you can do this then you don't delegate execution path derivation to the agent. The benefit is a predictable coherent world state where you understand the impact of { current state } x { action } without having to enumerate that huge cartesian product.by nostrebored
- This might be pretty big. One of my biggest frustrations with smaller models (especially MoE) is their failure to track workflow state at a high level. I'm constantly reminding them what we decided on or asking them to revisit, and reminding them eats context.
Seems like this might make that a lot less painful. And if not off the bat, with some minimal tuning or even just good prompting.
- I think open-ended simulation for agents will be a key component for training and planning. Similar as human dreams simulate different scenarios in our head. Biggest challenge will be simulating more abstract and complex systems.
Few months ago I did experiment with an open-ended world simulation for AI agent, where the simulated world was progressively building itself based on each of agent actions in open-ended manner. The idea was to give an agent infinite possibility regarding tool calling, where the tool call would be approved by the adjudicator, and the world state would change. The key issues with the PoC were:
Anyways the project came to be really funny when you watched agent struggling in desperation to perform real world actions which would be impossible in real world. Main observation was that when presented agent with current action budget, it modulated the creativity and how desperate its actions were.- World decoherence (tried to solve that with a poor graph implementation) - World flatness - high abstraction did not account for small events that would compound in real world - Start with empty context was real issue to get the agent to explore the world - Out of curiosity would you be willing to share the full system prompt for the agent in question described in this test?by walrus01
- Is there much evidence we use dreams to pre-emptively simulate scenarios?
Dreaming seems much more likely to be neurological tidying and emotional reprocessing. Helpful for identifying and surfacing long term subconscious needs but not for planning.
My dreams would be precisely useless for making plans from, unless those plans were to involve being caught in public wrapped only in a towel. And even then, I'm not sure they'd be particularly helpful.
by dofm - I agree; after running out of data on the internet, and humans being too slow to generate data, simulation is the only frontier left for improving things (training, datasets, reasoning). And it's probably the most ethical one too.
If nothing else I'm glad to see "world models" that are actually modeling some kind of worlds, instead of the term being applied as a hype layer for video/splats diffusion.
by avaer