Tracked topic
…Unlike reconstruction-based memory, the gate is trained directly against a closed-loop action-error signal . Its inference state is fixed at 4,224 bytes regardless of horizon, while a KV-cache…
…Probing which visual channels carry this signal reveals a sharp dissociation: removing color costs only 10.6pp of house identity accuracy, while removing texture costs 37.6pp, establishing texture and luminance as…
…We therefore propose Sign-Gated On-Policy Distillation (SG-OPD), which uses a binary verifier as a trust signal for the teacher at two complementary granularities: phased teacher sampling mixes in verifier…
…Rather than uniformly broadcasting sample-level advantages, GCPO assigns token-level advantages proportional to the difference between these contrastive predictions , allowing more precise and informative learning signals. Empirically, we find that GCPO…
…Using this signal, a Proposer model is trained via GRPO to continuously target the Generator's blind spots by synthesizing challenging queries and informative hints. The Generator is concurrently optimized via DPO…
…All learning derives from a single task-outcome signal. Its low-frequency trend credits selection and its high-frequency variation credits distillation. Experiments on ALFWorld and WebShop show that Skill1 outperforms prior…
…Crucially, we leverage these fine-grained attribution signals to guide downstream prompt optimization , establishing a closed-loop system that automatically corrects faults and boosts end-task performance by up to 7.62…
…Together these results signal that the agent- alignment lever is shifting from classical model alignment and prompt steering toward harness alignment and skill steering . View arXiv page View PDF Project page GitHub…
…While specialized Time Series Foundation Models (TSFMs) excel at forecasting based on numerical patterns, they remain unaware to real-world textual signals. Conversely, while LLMs are emerging as zero-shot forecasters, their…
…Building on this finding, we propose Probe&Prefill , which uses a lightweight linear probe to read the hidden-state signal and prefills the model's response with a steering sentence . Across all…