Peter Vamplew

I felt a great disturbance in the Force, as if thousands of Squaxx dek Thargo suddenly had exactly the same reaction @2000ad.bsky.social

When shaping your research agenda, your objective is to find the weirdest niche possible that still has the potential to change everything.

So, having not just one but two Australian Research Council DP EOIs rejected (one of which was led by a brilliant colleague of mine from UTS in Sydney), these are my thoughts.

Agreed. Also the quote marks around "the best" are critical. A colleague's proposal was ranked in the Top 10% of unsuccessful applications last year. This year the same application was placed in the Bottom 25%.

As Tina Turner used to sing "You're simply the beneficiary of a largely random process"

This is a hell of a coincidence. All morning I've had the song Ma Baker by Boney M stuck in my head. Now I check my bsky notifications, and someone called Ma Baker is following me?

Ex-pat Scot living in Australia. I had a vague recollection of you mentioning the Habs during the NHL season so I checked your profile.

By the way. @sharky6000.bsky.social if you're reading this thread of comments on my post, I want to make it 100% clear that I am insulting Hibs, not the Habs :-)

I've seen some papers that superficially look ok, but then you realise the Results are quite poor. That sounds like Hibs to me!

I have joint Australian and British citizenship. My Australian passport gets me into 182 countries without a visa. But soon that won't include the country where I was born.

Surely that should be ‘Barry’ ——————— ‘Hibs’

I haven't seen this previously with other IEEE journals, so I think it might be specific to this journal?

Exporting the US healthcare system might constitute an international act of aggression

I'm not a fan of some of terms used in this review form from IEEE Transactions on Emerging Topics in Computational Intelligence. Using words like "useless" and "dreadful" is unlikely to encourage the sort of constructive, respectful feedback that reviewers *should* be giving.

Glad to report that I submitted to both JAIR and TMLR this week (not the same paper) :-)

How long does it take to check that references don't correspond to reality? I just checked this chapter online, and it’s still available for sale with no indication that it is under investigation – not even a Notice of Concern. I also note that it has now been cited twice.

It's the 1 year anniversary of me reporting a case of fake references to Springer. It has a fictional reference to a paper I never wrote (#11) as well as multiple other AI hallucinations. So far Springer have done nothing beyond replying to me saying they are investigating.

lnkd.in/gR94VEMn

I once rode the dodgem cars at The Beach House in Glenelg with my kids The only other adult driver was Clive Owen who I think was with the young actor who played his son in The Boys are Back.

I'm guessing keyword mismatch - my PhD student Adrian Ly is looking at overestimated Q-values, but we hadn't found your work. I've alerted him to it now, so it appears social media engagement may be the solution!

On that note, please check out Adrian's work 😀
scholar.google.com/citations?hl...

This analysis of trends on arxiv shows the same as my previous evaluation of Google Scholar publications. Multi-objective reinforcement learning has seen huge growth as a research area in recent years. In particular 2025 saw a real spike in activity on this topic.

That's consistent with how they handled the 2023 case that I reported last year.

Last year I reported a 2023 NeurIPS paper for fabricated citations (about 9 cites out of 40+ in total). I still trust the core idea - in fact I intend to build on it. So that's an example of #2. But its still clearly misleading for that paper to remain available with those citation errors.

Actually I just checked and that paper has *not* been corrected. The NeurIPS proceedings still link to the original, uncorrected pdf.

In mid-2025 I reported fake references in a 2023 paper to NeurIPS Ethics and Grievances Committee. They said they had contacted the authors, who apologized & sent a corrected version. This story prompted me to check the NeurIPS proceedings, and it's still the original, uncorrected pdf.

Last year I reported a case of hallucinated references in a 2023 NeurIPS paper. After investigation, that paper was corrected and the NeurIPS Ethics and Grievances Committee told me that "In the future, NeurIPS plans to instate a systematic check of papers for reference related issues."

I think it was also Mike Brooks that first introduced me to Rodney's work in 1990, when Mike was still at Flinders.

If I were a foreign power doing "foreign interference" on a tiny fraction of Australian research projects, I wouldn't believe my luck that Australia's response was to extend grant time-scales to 16 months while failing to reverse its systematic decline in research funding.

Talk about bang-for-buck!

If the government wants research to work more with industry, they need to sort this out. It's impossible to plan anything with industry partners when you have to tell them its over a year to know whether a grant is successful or not.

Grrrrr...

If you're going to be using a Mac computer, keep a straightened out paper-clip nearby, just in case.