Hallucinations in peer review research are a current major problem. It signals a grave lack of care by researchers, and should trigger more scrutiny of the claims therein. VerusCite is a tool to extract and verify citations in academic articles. Try it out at veruscite-data.com
Andy Wheeler
@crimede-coder.bsky.social
Data scientist, with background in criminal justice. Consulting website at https://crimede-coder.com/ Personal blog at https://andrewpwheeler.com/
Hallucinations in peer review research are a current major problem. It signals a grave lack of care by researchers, and should trigger more scrutiny of the claims therein. VerusCite is a tool to extract and verify citations in academic articles. Try it out at veruscite-data.com
A UK think tank on the responsible use of AI in policing just put out a paper that has hallucinated references. See the citation report at veruscite-data.com/share/mTAk46...
New blog post out -- Notes on Document processing with LLMs. andrewpwheeler.com/2026/07/23/n... People should not be using the more expensive frontier models, there are cheaper alternatives for OCR/structured extraction.
Notes on document processing with LLMs
One of the recent papers that came across my X feed was a researcher digitizing a corpus of old Sears catalogs to estimate inflation more accurately: The technical skills to do this work are closel…
andrewpwheeler.com
New pod out with @crimede-coder.bsky.social! We talk about how technology can improve crime analysis, the future with AI, and the barriers to adoption. Terrific conversation with one of the smartest people I know. Spotify: open.spotify.com/episode/2hiB... Apple: podcasts.apple.com/us/podcast/u...
Using Technology to Improve Crime Analysis with Andy Wheeler
The Jeff-alytics Podcast · Episode
open.spotify.com
Citation verification results for “Reimagining-justice-peacemaking-criminology-and-alternative-justice-systems-in-the-Kenyan-context.pdf” — checked with VerusCite veruscite-data.com/share/tdvZvb...
Reimagining-justice-peacemaking-criminology-and-alternative-justice-systems-in-the-Kenyan-context.pdf — VerusCite results
32 citations · 31 verified · 0 hallucinations
veruscite-data.com
It is better to follow me for more updates at www.linkedin.com/company/crim... than here. But for a random tidbit, here is one of my favorite maps in my career. Proportion differences in pedestrian stops vs violent crime in NYC.
I have a new book out -- Large Language Models for Mortals: A Practical Guide for Analysts with Python This is an entry level book focused on using the foundation model APIs. Includes examples for OpenAI, Anthropic, Google, and AWS. crimede-coder.com/blogposts/20...
New pod out today with Charles Fain Lehman. We're talking about crime, disorder, why crime is falling and why people don't believe it. It's a great listen so check it out! Spotify - open.spotify.com/episode/4VWW... Apple - podcasts.apple.com/us/podcast/c...
I have an R WASM app up on my website. For experiments where you count things (e.g. reduce crime over time, increase arrest rates), this shows how long you need to evaluate to get a precise treatment effect estimate. crimede-coder.com/graphs/Plann...
xAI just released a voice cloning API. Very easy to use, I just created a few quick demo's (less than an hour total from recording script to text to speech). andrewpwheeler.com/2026/05/04/x...
xAI voice cloning API
xAI has just released an API to clone your voice. It is pretty simple, read a script, and then an API where you can have text to speech in that voice. Here is the python code after you have cloned …
andrewpwheeler.com
I am exploring offering short bootcamp tech classes, similar to an ICPSR summer course or what Paul Allison is offering at Horizons. Focus will be on technical skills that help you get a job in the private sector. Not help you write a paper. Let me know if interested! forms.gle/h2aBG1iRy9HY...
Short Course Interest
I am gathering interest in short, bootcamp style, technical courses (2 to 5 days maximum synchronous days). The focus of these courses will be on technical skills that help you get a job in the priv...
forms.gle
Repost of an ASEBP blog post I wrote -- there is no evidence that "discrimination" scorecards work, but there is decent evidence that simply having admin oversight of police use of force reduces officer involved shootings, taser uses, etc. crimede-coder.com/blogposts/20...
Evidence Based Oversight on Police Use of Force
What actually works to reduce officer use of force
crimede-coder.com
I am exploring offering short bootcamp tech classes, similar to an ICPSR summer course or what Paul Allison is offering at Horizons. Focus will be on technical skills that help you get a job in the private sector. Not help you write a paper. Let me know if interested! forms.gle/h2aBG1iRy9HY...
Short Course Interest
I am gathering interest in short, bootcamp style, technical courses (2 to 5 days maximum synchronous days). The focus of these courses will be on technical skills that help you get a job in the priv...
forms.gle
I made a single page (that I will update), with resources I have written for social science students/profs looking to transition into the private sector andrewpwheeler.com/job-advice-r... Majority of folks it is a minor effort to be able to compete for private sector "data" jobs.
Job Advice Resources
I have published many different resources for individuals to help transition from academia or public sector crime analysis roles to the private sector. This page is a catalog that I will continue t…
andrewpwheeler.com
Another experiment in LLM writing, I had gemini write a quick blog post on the tools to review papers, and how without any hard benchmarks it is just a market for lemons andrewpwheeler.com/2026/04/21/t... links to gemini chat session at end of post
The race to the bottom with AI tools
What we are seeing in the AI startup space is a perfect example of the “no moat” problem: if your core product is essentially just clever prompt engineering wrapped around someone else’s frontier m…
andrewpwheeler.com
One of the limitations of LLMs when asking for multiple potential outcomes, when just looking at the final produced text, they do not map to probabilities. You can look at the individual token probabilities though, and control false-positive/recall rates directly crimede-coder.com/blogposts/20...
Confidence in Classification using LLMS
Confidence in Classification using LLMS and Conformal Sets
crimede-coder.com
“Clinicians preserve uncertainty & iteratively refine differential diagnoses, whereas LLMs [across model generations]collapse prematurely onto single answers…. The risk is not just that LLMs are sometimes wrong but that their reasoning is brittle precisely where uncertainty & nuance matter most.”🧪🛟
New blog post, policing scholars should join the American Society of Evidence Based Policing, it is a much better org for collaborating with police than either ASC/ACJS. Renee Mitchell is the modern day Vollmer. crimede-coder.com/blogposts/20...
A tale of students taking easy online courses in lieu of more rigorous ones at Uni of Oregon, eighteenthelephant.com/2026/04/13/o...
Online courses, supply and demand, and academic integrity
What makes a college course popular or unpopular? I’ve long been interested in courses for non-science majors that satisfy “general education” requirements, their aim being to fos…
eighteenthelephant.com
New blog post, I show an example session of using the Gemini CLI to conduct crime analysis. Car breakins in Chapel Hill using public data. The current agent coding tools (not just Claude Code, also Codex and Gemini) are very good. crimede-coder.com/blogposts/20...
Working on something, if folks have examples where people had hallucinated citations (either journal articles or court proceedings is fine), appreciate send me the papers.
Did a coding session yesterday for an app idea I had. I have no clue how people are saying they are doing 10k lines of code a day in the coding apps (unless it is just crazy boilerplate). I did like 1k in 4 hours, but many iterations/tests.
Interview with Jason Elder on the LEA Podcast is out about my recent book www.leapodcasts.com/e/llms-for/ Check it out, and has the first promo code for the paperback version of the book
New blog post, Don't focus on the kids for gun violence prevention. The majority of gun offenders and victims are adults. crimede-coder.com/blogposts/20...
Don't Focus on the Kids
Do not focus on juveniles for violence reduction
crimede-coder.com
The Use-of-Force Continuum is intended to provide guidance for police officers on the appropriate level of force to use to gain compliance. Check the podcast (www.buzzsprout.com/2413505/epis...) where Dr. Scott Mourtgos discusses the "muddy middle" of the continuum.
Police Use-of-Force Continuum - Police In-Service Training
The Use-of-Force Continuum is intended to provide guidance for police officers on the appropriate, proportional level of force to use when trying to gain compliance. There seems to be a logical ...
buzzsprout.com
Professors in social science should be teaching students Python, not R. This blog post goes over a sample on job postings on LinkedIn for data science to illustrate the disparity. andrewpwheeler.com/2026/03/22/s...
Stop Teaching R. Teach Python.
There has been a slight transition in social science teaching since I have been a student and professor over the past ~15+ years. In the aughts, it was still common to teach students in legacy, clo…
andrewpwheeler.com
New blog post on how I use Claude Code to help me write. Using it for copy-editing should not be controversial at all. But if you want to use it for writing, have it review your prior writing before generating new content based on an outline. andrewpwheeler.com/2026/03/20/u...
Using Claude Code to help me write
Using LLMs to help you write is understandably a touchy subject for many. There is quite a bit of AI slop coming out now, as it is really easy to just have the LLM tools think for you and write sup…
andrewpwheeler.com
New blog post, comparing local models vs API usage. Even if you have a GPU where you can run a model locally, it often does not make sense given the low cost of using APIs. crimede-coder.com/blogposts/20...
Using local LLM models vs APIs
Comparing cost and throughput for local LLM models vs API
crimede-coder.com
I have a few different promo codes floating around the for the book. LLMDEVS gets you $25 off of the epub. TWOFOR1 gets you $30 off if you purchase multiple books. crimede-coder.com/store