Capabilities · Concept
Diffusion Language Models (DLMs): Systematic Evaluation
Non-autoregressive diffusion LMs positioned as reasoning substrates improved by causal latent revision methods such as CaLR.
Forecast horizon
Connections
Connections · 3
How this node ties into the rest of the map, and the evidence behind each link.
from · standardizes 4/10
CaRE: Compute-Aware Remasking Evaluation for MDLMs CaRE audits remasking strategies for masked diffusion LMs under matched compute, metrics, and stochasticity.
+4 growth to · enables 3/10
AI Agents Reshaping Knowledge Work Diffusion language models with any-order infill capability enable new interactive drafting workflows for knowledge workers such as radiologists.
+3 growth from · applies to 3/10
DLMRec: Diffusion Language Model for Recommendation DLMRec specializes discrete diffusion LMs to generative recommendation instead of autoregressive next-token recommenders.
+3 growthSignal sources
Signal sources
Dated facts from primary sources in this direction.
Task horizon doubling Mar 2025
The length of software tasks AI agents can do autonomously at 50% reliability has doubled about every 7 months — and since 2024 closer to every ~3 months.
METR → Benchmarks saturating Apr 2025
In one year scores rose by 18.8, 48.9 and 67.3 points on MMMU, GPQA and SWE-bench; real-world software solve rate jumped from 4.4% to 71.7%.
Stanford HAI — AI Index 2025 → Autonomous coding 2025–2026
On SWE-bench Verified (500 real GitHub issues), autonomous coding agents reached ~80–86% by late 2025, up from under 50% in early 2025.
Epoch AI →