Line of inquiry
Inquiring lines›How can we train models to be prop…›What training approaches and contr…›this line of inquiry
Does reinforcement learning create fundamentally new reasoning or optimize existing capabilities?
A broader line of inquiry — a family of 67 specific questions the research asks around this. Follow one into its inquiring-line page, or move sideways to a related line below.
Questions in this line of inquiry 67
Specific inquiring lines the field asks around this — ordered from the most general framing down to the most specific angle.
- Can RL create new reasoning primitives that pretraining never established?
- Can reinforcement learning add new capabilities or only remove inaccurate knowledge?
- Does RL teach models new reasoning or just better timing?
- Does RL refine existing knowledge or discover entirely new capabilities?
- Does RL teach models when to use reasoning or how to reason?
- Does RL amplify existing reasoning or create genuinely new computational strategies?
- Does RL primarily teach when to use reasoning or how to reason?
- Can extended RL training unlock genuinely new reasoning strategies models cannot discover otherwise?
- How does RL refine reasoning paths without simply adding model capability?
- When does reinforcement learning actually produce true reasoning gains in models?
- Does targeting the edge of competence during RL pretraining unlock true reasoning gains?
- Can base models spontaneously produce reasoning traces without any RL training?
- Does RL primarily teach output format rather than new knowledge?
- Can RL teach when to use reasoning versus when to respond directly?
- Can RL training teach models when to activate reasoning versus when to skip it?
- Does reinforcement learning require sufficient pretraining to be effective?
- Why do reasoning gains from RL require models trained with headroom and edge-of-competence data?
- What limits RL's ability to scale for reasoning at training time?
- What makes pretraining composition more important than reward engineering?
- Can reinforcement learning close the gap between LLM reasoning and action?
- Can reinforcement learning add missing domain knowledge to fine-tuned reasoning models?
- Can one training example activate mathematical reasoning in RL-trained models?
- How does pretraining quality versus quantity affect downstream RL gains?
- What makes reasoning capability a pre-training rather than post-training phenomenon?
- Why do overtrained domains show different RL training outcomes than novel tasks?
- How does pretraining determine what RL can later teach a model?
- Can pretraining signals unlock latent reasoning that post-training merely activates?
- Does reinforcement learning teach models how to reason or when to reason?
- How do extrapolative and contextual generalization measure RL reasoning gains?
- How do two-phase training dynamics explain reasoning emergence?
- What training duration is actually needed for RL to expand capabilities?
- How do self-evolving curricula help RL break beyond base model capability boundaries?
- What capacity threshold determines whether RL teaches activation versus shortcut learning?
- Why does the pretrained prior determine the exploration ceiling?
- What distinguishes RL that creates new capabilities from RL that merely teaches timing?
- Does the pretrained model prior limit RL search capability more than the optimization algorithm itself?
- How does baseline capability level affect RL improvement ceiling?
- Does RL training activate latent meta-learning capacity or create it from scratch?
- What does RL post-training actually teach reasoning systems?
- Can safety training reduce or eliminate metagaming after capabilities training?
- Why does extended reasoning training improve exploration without adding new capabilities?
- How does RPT compare to learning when versus how to deploy reasoning?
- Can the exploration ceiling be raised beyond what pretraining established?
- What scaling properties emerge from RL training dynamics beyond verification?
- How does model scale affect anticipatory behavior in structured training?
- Does RL training redirect self-doubt into productive gap analysis?
- What pretraining formats encode latent reasoning strategies that RLVR can surface?
- How does the pretrained prior constrain the ceiling for empathy RL improvements?
- Does RL training actually restore the critical thinking that reasoning models lose?
- Can reinforcement learning resolve the gap between knowing facts and knowing how to act?
- Do emergent abilities result from genuine new capabilities or implicit in-context learning?
- What training interventions could close the perception-action gap?
- Why does pre-training provide the raw material for emergent thinking?
- Why does early experience provide better warm-starts for downstream reinforcement learning?
- Can one training example activate mathematical reasoning without reinforcement learning?
- How do RL training and base models differ in creating MI peaks?
- How much does pretraining contribute to ToM performance versus task-specific training?
- How does early branch divergence differ from late branch divergence in supervision signals?
- Which recipe choices determine the asymptotic ceiling in RL training?
- What makes a task at the edge of competence optimal for RL?
- What role does pretraining play in distinguishing system capability from deployed behavior?
- What structural differences emerge between early generic skills and later meta-strategy skills?
- Can models be trained to verify feasibility before proposing plans?
- Can explicit goal state scaffolding at inference time transfer to autonomous tracking through training?
- How does sliding the start state backward create informative learning signals?
- Can models generate their own training curriculum during offline dreaming?
- How can weak-to-strong progressive training target planning without interfering with grounding?