🛰️ Daily AI Frontier
‹ back to 2026-08-08

CalibForge: Adversarial Solver Calibration for Scaling Learnable Terminal Tasks

Research LLM Agents

Ranking

Overall 77
Content 80
Popularity 69

Observed public metrics from 1 member.

Merged summary

TL;DR - CalibForge is an autonomous system that synthesizes terminal-agent training tasks and calibrates their difficulty against verified solver behavior, producing 5,431 tasks that yield large gains on agentic coding benchmarks. It matters because it targets task learnability relative to a solver, not just executability, as a data-construction principle.

  • Two calibration strategies define a "solver-relative learnable zone": multi-solver calibration exploits disagreement across a heterogeneous solver pool, while contrastive calibration enforces a strong-pass/weak-fail relation, both anchored in demonstrated solvability.
  • Ablations report both strategies outperform authoring-plus-executable-validation alone and ordinary single-solver feedback.
  • Models trained on the full 5,431-task collection reach 32.58% and 47.57% on Terminal-Bench 2.0.
  • Largest reported gains over the corresponding base model: +24.71 points on Terminal-Bench 2.0, +27.68 on SWE-bench Pro, and +30.04 on Doc2Repo, suggesting transfer beyond the terminal domain.

Sources (1)

CalibForge: Adversarial Solver Calibration for Scaling Learnable Terminal Tasks

arXiv cs.LG Fanzhe Meng, Guoxin Chen, Jiale Zhao, Shuang Sun, Zhiyu Lin, Wayne Xin Zhao, Ruihua Song, Ji-Rong Wen, Kai Jia 2026-08-06 arXiv:2608.06352
Public signals Hugging Face upvotes 23
Providers: Hugging Face · Upvotes 23 OpenAlex · N/A Publisher · N/A Semantic Scholar · N/A X · N/A Fetched 2026-09-07 14:27:46.876664 UTC

TL;DR - CalibForge is an autonomous system that synthesizes terminal-agent training tasks and calibrates their difficulty against verified solver behavior, producing 5,431 tasks that yield large gains on agentic coding benchmarks. It matters because it targets task learnability relative to a solver, not just executability, as a data-construction principle.

  • Two calibration strategies define a "solver-relative learnable zone": multi-solver calibration exploits disagreement across a heterogeneous solver pool, while contrastive calibration enforces a strong-pass/weak-fail relation, both anchored in demonstrated solvability.
  • Ablations report both strategies outperform authoring-plus-executable-validation alone and ordinary single-solver feedback.
  • Models trained on the full 5,431-task collection reach 32.58% and 47.57% on Terminal-Bench 2.0.
  • Largest reported gains over the corresponding base model: +24.71 points on Terminal-Bench 2.0, +27.68 on SWE-bench Pro, and +30.04 on Doc2Repo, suggesting transfer beyond the terminal domain.
item →