From Pareto to Preference: Personalized Test-Time Scaling via Amortized Agentic Policy Discovery
To reduce the overhead of repeated policy discovery for new user profiles, we propose PersonTTS, an amortized agentic policy-discovery framework that reuses prior search experience through requirement-matched controller initialization and source-distilled procedural guidance, while retaining target-profile evaluation for every candidate.
Key points
- Test-time scaling (TTS) improves the reasoning capabilities of large language models by allocating additional inference computation.
- Existing approaches to improving TTS efficiency largely optimize accuracy against one resource dimension at a time, advancing either the accuracy--cost or accuracy--latency Pareto frontier.
- We formulate Personalized Test-Time Scaling as discovering executable controllers that maximize the joint satisfaction rate of user-specific requirements.
- Experiments on AIME and HMMT show that PersonTTS substantially outperforms strong TTS baselines in joint requirement satisfaction on unseen user profiles and held-out problems.
Sources (1)
- [1]From Pareto to Preference: Personalized Test-Time Scaling via Amortized Agentic Policy DiscoveryarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 7, 08:46 AM
To reduce the overhead of repeated policy discovery for new user profiles, we propose PersonTTS, an amortized agentic policy-discovery framework that reuses prior search experience through requirement-matched controller initialization and source-distilled procedural guidance, while retaining target-profile evaluation for every candidate.
Test-time scaling (TTS) improves the reasoning capabilities of large language models by allocating additional inference computation.
Extractive summary: sentences quoted from the sources.