> We propose Recursive Language Models (RLMs), a general inference paradigm that treats long prompts as part of an external environment and allows the LLM to programmatically examine, decompose, and recursively call itself over snippets of the prompt.
> We find that RLMs can successfully process inputs up to two orders of magnitude beyond model context windows and, even for shorter prompts, dramatically outperform the quality of vanilla frontier LLMs and common long-context and coding scaffolds [...] across four diverse long-context tasks while having comparable cost.
I had a similar thought the other day. When doing a research task, you don't want to crap up the context with all the web scrapes. But you want to ask follow up questions on the full context, not the anemic subagent summaries. So what you actually want is an "extended context" you can grep.
funfunfunction 6 hours ago [-]
Cool project. A team at work was building something similar to internal use.
I'm curious how this compares to just using Claude Code directly and giving it a dump of the agent traces? It seems like Claude could probably do some of the same diagnostics / trace grouping to identify failure patterns. Why use a custom harness?
mikepollard_dev 6 hours ago [-]
Yeah, fair question. For a small number of traces just dumping them into Claude Code can work well.
However, once you're at production scale the problem changes. You can't always fit 10,000+ traces in Claude Code and still have it be effective especially when the relevant pattern of agent failures may only become apparent when you pass that many in. That's where the RLM based methodology helps. HALO recursively decomposes the trace data into smaller investigations, analyzes those sub-pieces, and then synthesizes those up to determine the recurring harness-level failure modes better than Claude Code or Codex ever could at a large scale.
ilusion 2 hours ago [-]
I'm very curious to see a benchmark for this - have toyed with the idea myself but haven't put in the hard work to test these hypothesis on extracting learning signal from deep-agent traces.
funfunfunction 2 hours ago [-]
There's some benchmarks in the repo for AppWorld. Looks promising
https://github.com/alexzhang13/rlm
> We propose Recursive Language Models (RLMs), a general inference paradigm that treats long prompts as part of an external environment and allows the LLM to programmatically examine, decompose, and recursively call itself over snippets of the prompt.
> We find that RLMs can successfully process inputs up to two orders of magnitude beyond model context windows and, even for shorter prompts, dramatically outperform the quality of vanilla frontier LLMs and common long-context and coding scaffolds [...] across four diverse long-context tasks while having comparable cost.
https://arxiv.org/abs/2512.24601
I had a similar thought the other day. When doing a research task, you don't want to crap up the context with all the web scrapes. But you want to ask follow up questions on the full context, not the anemic subagent summaries. So what you actually want is an "extended context" you can grep.
I'm curious how this compares to just using Claude Code directly and giving it a dump of the agent traces? It seems like Claude could probably do some of the same diagnostics / trace grouping to identify failure patterns. Why use a custom harness?
However, once you're at production scale the problem changes. You can't always fit 10,000+ traces in Claude Code and still have it be effective especially when the relevant pattern of agent failures may only become apparent when you pass that many in. That's where the RLM based methodology helps. HALO recursively decomposes the trace data into smaller investigations, analyzes those sub-pieces, and then synthesizes those up to determine the recurring harness-level failure modes better than Claude Code or Codex ever could at a large scale.