It's an old paper (from 2025, so, a decade ago in AI years), but the concept is still fascinating. And I'm yet to see it show up in any production models.

If multiple models can use cache representations for this kind of enrichment, the KV cache representations of different models must be somewhat compatible.

What stops us then from going a step further, and producing a model family where all models are "KV aligned", and each model can utilize the KV cache of other models directly?

So, an "expensive" reasoning model can use its full faculties to plan, but "delegate" simple subgoals to a smaller model. That smaller model can access the large model's intent directly, as rich KV cache representations - with no prefill recompute and no associated "handover" latency. Or, likewise, a "cheap" small model can generate a diminished but highly compact KV cache that the "expensive" model can then operate on - for example, for skimming a large file for shallow patterns.

A few months ago I asked why semantic representation rather than text wasn't used, since natural language seems quite a lossy representation for semantic concepts:

https://news.ycombinator.com/item?id=47195212

I wouldn't have thought to use it for LLM-to-LLM communication, though

I feel like multimodal models that can read images should work differently than they do. My understanding is that multimodal models basically first generate an image embedding and then the model is trained to interpret that embedding, but in the same way that text is lossy, it seems like the embedding would be as well. Why don't multimodal models learn to interpret images themselves without an embedding? Or e.g., by passing some "prompt" to the embedding model?

So the models will not only be using more and more Neuralese in their CoT (like GPT-6), but different agents will also be able to communicate with each other in Neuralese. It's not looking good for monitorability.

[flagged]

[flagged]

[dead]

[dead]