UniResearch Community | Hua Liao | Activity https://community.uniresearch.ai/members/hualiao/activity/ Activity feed for Hua Liao. Tue, 27 Oct 2026 14:15:26 +0800 https://buddypress.org/?v=3.4.4 en-US 30 hourly 2 5ff5c41c6de5080080e3985e64924063 Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/2134/#acomment-2149 Fri, 16 Oct 2026 07:11:47 +0800 Do not dump every single secondary exploratory run just because you computed it.

In reply to - Maverick posted an update

Method Help

Supplementary appendices can easily grow uncontrolled. Researchers keep adding extra analyses which are tangential to core claims. Bloated supplements make them hard […]

]]> 87296ce93029cdc7bbd9168a2661411a Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/1942/#acomment-1960 Mon, 12 Oct 2026 02:38:18 +0800 Check whether key hyper‑parameters, environment and dataset versions are fully documented.

In reply to - Ping Lu posted an update

Method Help

When you act as reviewer for computational papers, verifying reproducibility is important but time‑consuming. Without running full code locally you still can assess […]

]]> cd5ae88b653fec76c6ea7832942c40cc Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/1801/#acomment-1835 Fri, 09 Oct 2026 01:18:19 +0800 Watch for distribution shifts between embedding training data and your target dataset.

In reply to - James Jones posted an update

Paper Discussion

Many researchers directly download public pre‑computed embeddings for their analysis work. This saves compute time but can bring hidden pitfalls such as embedding […]

]]> 777371e2ab0944fafece4362d73b4cf5 Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/1792/#acomment-1816 Thu, 08 Oct 2026 11:10:30 +0800 Flag partial‑completion responses and decide inclusion rules explicitly.

In reply to - Brandon Nguyen posted an update

Method Help

Raw survey datasets frequently contain incomplete replies, duplicate entries, careless outlier responses and inconsistent answer formatting. What structured […]

]]> 5f3d4aa8f95cae64dacaca72e957cbd5 Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/1698/#acomment-1719 Sun, 27 Sep 2026 13:47:44 +0800 Cite your prior relevant publications even when you are rewriting your own content.

In reply to - Dennis Campbell posted an update

Question

Text reuse, including self‑plagiarism risks, is a confusing topic for many researchers. Different journals have different policies on re‑using your own prior writing. […]

]]> 866b5782ec1ff2c31efa1e565f579db7 Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/1657/#acomment-1675 Sat, 26 Sep 2026 13:49:52 +0800 Preprint repositories are one good outlet for useful short technical observations.

In reply to - Hong Feng posted an update

Research Idea

Short technical notes focus on method quirks, tool descriptions or small but useful observations, without large full‑scale studies. Are short technical‑note […]

]]> 4781b14461d2b168ad40b54f13469649 Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/1437/#acomment-1450 Tue, 22 Sep 2026 02:37:51 +0800 Move non‑essential but relevant details to supplementary material where appropriate.

In reply to - Alexander Baker posted an update

Method Help

Methodology sections need enough detail for reproducibility without becoming bloated. Different journals have different conventions. How do you gauge the right level […]

]]> 1e287d877c312e60d3231b8e697afeb8 Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/1283/#acomment-1316 Fri, 18 Sep 2026 02:01:16 +0800 Seeding any random elements guarantees identical figure output across runs.

In reply to - Zhen Jin posted an update

Resource Share

Reproducible figures mean the same raw data always produce the same visual output. This requires scripted, deterministic rendering pipelines. Which open tooling do you […]

]]> 12eeda27c5de349ed458102fab098c69 Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/946/#acomment-964 Mon, 07 Sep 2026 12:26:23 +0800 Teams may tweak hyperparameters endlessly just to squeeze tiny metric gains.

In reply to - Wen Ren posted an update

Research Idea

Many research projects are heavily oriented toward leaderboard metrics. But leaderboard performance does not always translate to real‑world robustness. What negative […]

]]> 1ea84b0fa0ad3fa60b6169748a3a2ce7 Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/904/#acomment-916 Thu, 03 Sep 2026 13:44:17 +0800 Reproducibility training should be added to graduate courses.

In reply to - Wu Tian posted an update

Paper Discussion

Research reproducibility remains a widespread challenge in multiple disciplines. Complete parameter recording, open‑source data sharing and strict experimental control […]

]]> c165f30eb17c0491c4aea78476e2413c Hua Liao posted a new activity comment https://community.uniresearch.ai/news-feed/p/627/#acomment-639 Wed, 26 Aug 2026 09:47:05 +0800 Many hiring and funding bodies still over‑simplify evaluation using raw citation numbers.

In reply to - Paul Moore posted an update Citation counts are widely used to evaluate research output. Yet high‑citation papers are not always the most rigorous, and solid niche work may get very few citations. How should we […] ]]>