Tootfinder

Learning from Trials and Errors: Reflective Test-Time Planning for Embodied LLMs
Yining Hong, Huang Huang, Manling Li, Li Fei-Fei, Jiajun Wu, Yejin Choi
https://arxiv.org/abs/2602.21198 https://arxiv.org/pdf/2602.21198 https://arxiv.org/html/2602.21198
arXiv:2602.21198v1 Announce Type: new
Abstract: Embodied LLMs endow robots with high-level task reasoning, but they cannot reflect on what went wrong or why, turning deployment into a sequence of independent trials where mistakes repeat rather than accumulate into experience. Drawing upon human reflective practitioners, we introduce Reflective Test-Time Planning, which integrates two modes of reflection: \textit{reflection-in-action}, where the agent uses test-time scaling to generate and score multiple candidate actions using internal reflections before execution; and \textit{reflection-on-action}, which uses test-time training to update both its internal reflection model and its action policy based on external reflections after execution. We also include retrospective reflection, allowing the agent to re-evaluate earlier decisions and perform model updates with hindsight for proper long-horizon credit assignment. Experiments on our newly-designed Long-Horizon Household benchmark and MuJoCo Cupboard Fitting benchmark show significant gains over baseline models, with ablative studies validating the complementary roles of reflection-in-action and reflection-on-action. Qualitative analyses, including real-robot trials, highlight behavioral correction through reflection.
toXiv_bot_toot

Trump’s pledge to raise the global baseline tariff rate from 10% to 15% is reverberating across markets, with investors now weighing whether a recent rally in overseas equities can hold.
In a post on TruthSocial, Trump said he would immediately increase the worldwide tariff rate to 15% following a Supreme Court decision related to trade authority. He also signaled that additional, “legally permissible” tariffs could be introduced in the coming months, reinforcing a more aggressive trade…

Kolanovic warns of reversal in overseas stocks after Trump signals tariff hike — Seeking Alpha
President Donald Trump’s pledge to raise the global baseline tariff rate from 10% to 15% is reverberating across markets, with investors now weighing whether a recent rally in overseas equities can hold.

Apropos of this second-latest comment on this fascinating thread, if your argument boils down to stuff that was removed from a draft spec three years ago, then your argument might be moot:
https://github.com/w3c/wcag3/issues/642#issuecomment-4360244086
Because…

要不是这周的《字谈字畅》我都不知道 text-autospace: normal 已经 Baseline 2025 Newly Available 了。可惜 text-autospace: replace 还没有浏览器在做。

#CSS

The Type — 文字 / 设计 / 文化 — 字谈字畅 276：「它可以拓宽你的认知边界」

Compressing Transformer Language Models via Matrix Product Operator Decomposition: A Case Study on PicoGPT
Younes Javanmard, Tanmoy Pandit, Masoud Mardani
https://arxiv.org/abs/2603.28534 https://arxiv.org/pdf/2603.28534 https://arxiv.org/html/2603.28534
arXiv:2603.28534v1 Announce Type: new
Abstract: Transformer-based language models achieve strong performance across NLP tasks, but their quadratic parameter scaling with hidden dimension makes deployment on resource-constrained hardware expensive. We study Matrix Product Operator (MPO) decomposition as a principled compression method for transformers. MPO factorises weight matrices into chains of low-rank cores, with approximation quality controlled by the bond dimension chi. We replace every nn.Linear layer in PicoGPT, a GPT-2-style character-level language model with about 1M parameters, with an MPOLinear module parameterised as an MPO chain. Cores are initialised either by TT-SVD from pretrained dense weights or from random initialisation, and trained using standard PyTorch autograd without a custom backward pass. We derive balanced factorisation schemes for the five distinct weight shapes in PicoGPT and evaluate bond dimensions chi in {4, 8, 16, 32} on Tiny Shakespeare. MPO compression achieves up to 13x compression per transformer block at chi = 4. At chi = 16, the model uses 191,872 parameters instead of 1,020,224 while retaining 97.7% of baseline token accuracy (51.6% vs 52.8%). Reconstruction error follows the expected trend and is lower for three-site than two-site factorisations at the same bond dimension. The chi = 8 model gives the best accuracy per parameter, exceeding the dense baseline by 2.7x on this metric. These results show that MPO parameterisation is a practical and theoretically grounded alternative to low-rank methods and unstructured pruning for transformer compression.
toXiv_bot_toot

SELAUR: Self Evolving LLM Agent via Uncertainty-aware Rewards
Dengjia Zhang, Xiaoou Liu, Lu Cheng, Yaqing Wang, Kenton Murray, Hua Wei
https://arxiv.org/abs/2602.21158 https://arxiv.org/pdf/2602.21158 https://arxiv.org/html/2602.21158
arXiv:2602.21158v1 Announce Type: new
Abstract: Large language models (LLMs) are increasingly deployed as multi-step decision-making agents, where effective reward design is essential for guiding learning. Although recent work explores various forms of reward shaping and step-level credit assignment, a key signal remains largely overlooked: the intrinsic uncertainty of LLMs. Uncertainty reflects model confidence, reveals where exploration is needed, and offers valuable learning cues even in failed trajectories. We introduce SELAUR: Self Evolving LLM Agent via Uncertainty-aware Rewards, a reinforcement learning framework that incorporates uncertainty directly into the reward design. SELAUR integrates entropy-, least-confidence-, and margin-based metrics into a combined token-level uncertainty estimate, providing dense confidence-aligned supervision, and employs a failure-aware reward reshaping mechanism that injects these uncertainty signals into step- and trajectory-level rewards to improve exploration efficiency and learning stability. Experiments on two benchmarks, ALFWorld and WebShop, show that our method consistently improves success rates over strong baselines. Ablation studies further demonstrate how uncertainty signals enhance exploration and robustness.
toXiv_bot_toot