Here is a recording of my talk @CogCompNeuro last week on quantifying similarity in neural recordings Very much enjoyed the Keynote Tutorial format for the conference. Links to our tutorial notebook below 👇 youtube.com/live/-xmLMWC2YlI…

Great blog post from @ItsNeuronal on the interpretation of Log-Likelihood Ratios for model comparison, including a (new to me!) result on how to obtain a significance threshold for declaring one model better than another: neurostatsblog.github.io/202…

Neural dynamics are complex, varying in time and across experimental conditions. 🎉 Excited to share our #NeurIPS2025 work: we introduce Conditionally Linear Dynamical Systems (CLDS) as an interpretable and flexible model of neural dynamics. 🧵[1/9]

Congrats to our very own Ipshita Zutshi on being awarded the Peter & Patricia Gruber International Research Award in Neuroscience!! #SfN25

Senator Schumer is no longer effective and should be replaced. If you can’t lead the fight to stop healthcare premiums from skyrocketing for Americans, what will you fight for?

The application for a research fellowship at the Flatiron Institute in the Center for Computational Math is now live! This includes positions for ML and stats. The deadline is Dec 1. Links below with more details.

⏳ Deadline is next Friday! Apply now to join our #SfN25 workshop on pynapple & nemos. Learn to analyze & model neural data using software developed at @FlatironInst. Accommodations meals included! 👉 neurorse.flatironinstitute.o…

We don't know how good we have it, in historical terms. That means both that we're closer to a better country than we feel, and also that things could get worse than we appreciate.

Excited to announce that our paper on "Comparing noisy neural population dynamics using optimal transport distances" has been selected for an oral presentation in #ICLR2025 (1.8% top papers). Check the thread for paper details (0/n). Presentation info: iclr.cc/virtual/2025/oral/31….

Excited to share our JOSS paper (joss.theoj.org/papers/10.211…) and code (github.com/probml/dynamax)! Special thanks to our reviewers — Dynamax v1.0 is much improved thanks to their feedback!

Thousands of UAW members are international workers who conduct critical research to advance lifesaving treatments for diseases like cancer, Alzheimer’s, and more. Hear from international scholars across the country why they're taking action on 4/8 to #KillTheCuts.

UAW Region 4 CAP mobilization happening TODAY: members across the region are calling key Senators and Congresspeople to demand "No Cuts to NIH Funding!" An attack on one of our members is an attack on all of us!

Thank you to everyone who joined us for our first and second Pynapple & @nemos_neuro workshops of the year at @FlatironCCN , NYC! Stay tuned for future event announcements. #EastCoast

To save some suspense, industry is not doing a lot of curiosity driven fundamental research.

🧵 Many “hot takes” about the reduction of @NIH indirect grant costs to 15%. Let’s lay out the facts about the university grant management process & accounting. And then use this to project likely short and longer term impacts of the policy.

It seems few people know what an “indirect cost” is or why it has to be 40-60%. The reason the government forced universities to raise their indirect costs up to (typically) 40-60% was to force a huge amount of regulations on the universities while also minimizing the bookkeeping to comply with those regulations. This includes the work by contract managers, compliance lawyers, accountants, safety management, etc., who are required by the government per the terms of the contract. If universities had to allocate all those categories of labor to each contract hour-by-hour it would require too much bookkeeping, which would waste money. (I’m setting aside for now the question of whether or how much the regulations are wasting money and only discussing how you bookkeep the effort to comply with the regulations.) So to save money, while also requiring universities to do these types of work, the government requires universities to roll those categories of labor into “cost pools” that must be allocated as a percent of the technical work in each of the contracts. While the actual “overhead” might be only 15%, these pooled labor charges that are required by the government are typically much more. Second, the government doesn’t allow the universities to figure out their own indirect rates. These rates are determined by the federal government through audits every couple of years. The government then sends a document telling the university what rate to use for its cost pools. For example, the University of Colorado was told by the DHHS to use 54% (colorado.edu/controller/site…) and U. Nebraska was told by DHHS to use 55.5% (uofnelincoln.sharepoint.com/…). 40-60% is not only reasonable to fulfill the terms of the contract, it is the rate that the government tells the university it can charge for all the work the government requires the university to do. So if the government wants to reduce the indirect rate to 15%, then it needs to do one of these two things: Either (A) eliminate all the federal regulations that force the universities to do those categories of work (compliance, accounting, management, safety management, tracking harmful chemicals, etc.) Or, (B) stop requiring universities to pool those real costs into the “indirect cost” category and allow universities to include them in the “direct costs” of the contract. If the government chooses (A), then the safety rails have been entirely removed. (Even if the government lowers the regulations without entirely eliminating them, the costs they impose will still be real costs that probably come out to more than 15%.) Or, if it chooses (B), then the direct costs will go way up and research will actually be less efficient because all the bookkeeping, not more efficient. But if the government caps the indirect rate at 15% without doing either (A) or (B), then it will be impossible to do research for the federal government without going bankrupt. That’s the worst possible choice. It will kill research in the US. Is that what we want? I can explain it for you but I can’t understand it for you. It’s up to the reader not to be ignorant.

RT @BrainsExplained: Clearly Ivy Leagues are used here as examples to rile up conservatives, but rates are higher at places like Seattle Ch…

Imagine if academia were given 500B for AI research... it would be absolutely revolutionary compared to one company stockpiling GPUs. That's like 100 CMUs. For 0.01% of that money the right lab could profoundly advance the field.

I love assuming things are accurate. It has never made me look foolish even once.

16/n My lab at @NYU_CNS will combine the systems neuroscience and cell signaling to understand how motivational dynamics are computed biochemically in intact neural circuits, in real time, and during behavior. If this excites you too, please reach out! stephenzhanglab.org/