Temporal abstraction alleviates spectral mismatch in FB representations
What happened
Seyed Mahdi B. Azad and co-authors propose using temporal abstraction to alleviate spectral mismatch in forward-backward (FB) representations; the paper has been accepted to NeurIPS 2026. The authors state that a spectral mismatch exists between the high-rank transition dynamics of continuous environments and the low-rank bottleneck, and that temporal abstraction acts as a low-pass filter, suppressing high-frequency components and reducing the effective rank of the successor representation while preserving a formal bound on value function error. Experiments show that this spectral alignment is a key factor for stable FB learning under high discount factors; the authors state that bootstrapping is prone to error in this setting.
AI 根据报道生成 · 2 days ago更新
Reporting timeline
Follow the reporting to explore different perspectives.
- arXiv cs.ROSpectral Alignment of Forward-Backward Representations: Mitigating Spectral Mismatch via Temporal Abstraction
The paper analyzes the spectral mismatch between the high-rank transition dynamics of continuous environments and the low-rank bottleneck in forward-backward (FB) representations, and proposes temporal abstraction to mitigate this problem. Starting from the spectral properties of the transition operator, the authors show that temporal abstraction acts like a low-pass filter: it suppresses high-frequency components and reduces the effective rank of the successor representation, while preserving a formal bound on the value function error. Experiments show that this alignment is a key factor for stable FB learning at high discount factors. The work has been accepted to NeurIPS 2026.
Activity for this event
- Current
- 3
- Peak
- 10Nov 9 14:00
- Past 24 hours
- -50%
The trend compares participants observed throughout the same period, so its range may be smaller than the current interest count. Move or tap to inspect each hour; use the arrow keys on a keyboard.