Tootfinder

Dallas Cowboys Coaching Staff: The Catalyst for Success in 2025 https://insidethestar.com/dallas-cowboys-coaching-staff-the-catalyst-for-success-in-2025

Ethiopians Brew Success as Coffee and Cash Pile Up Thanks to Transformational Sustainable Forestry Program https://www.goodnewsnetwork.org/ethiopians-brew-success-as-coffee-and-cash-pile-up-thanks-to-sustainable-…

Ethiopians Brew Success as Coffee and Cash Pile Up Thanks to Transformational Sustainable Forestry Program
Coffee is a major lifeblood of Ethiopia's economy, (a quarter of the whole) accounting for around half of the livelihood of 15 million people

Travis Kelce opens up about returning to Chiefs, spurning retirement: 'Last year wasn't a success for me'

https://www.cbssports.com/nfl/n…

Pro Ball Buddy Mindset | Mental Strategies For Basketball Success With Teige Morrell
Great Australian Pods Podcast Directory: https://www.greataustralianpods.com/pro-ball-buddy-mindset-mental-strategies-for-basketball-…

It all fits. Suppose you’re corporate leadership. You’re laying people off / refusing to hire because you have no clue what things will look like even 1 or 2 years in the future, and you are •terrified•.
Meanwhile, your investors, your gullible trend-chasing investors, are already mentally dividing companies into “pre-AI” and “post-AI.” And not only do you have no idea how to be “post-AI,” but you have no idea whether you’ll be solvent next year. What do you tell them?!??
Obviously you tell them you’re not hiring because you’re having such incredible success with AI! Maybe even you’re going to lay off all the senior people! They’ll love that!
Now you’re not failing and flailing — you’re leading the march into the future! Yay!

The full formula for the probability of "success" is:
p = {
1/(2^(-n 1)) if n is negative, or
1 - (1/(2^(n 1))) if n is zero or positive
}
(Both branches have the same value when n is 0, so the behavior is smooth around the origin.)
How can we tweak this?
First, we can introduce fixed success and/or failure chances unaffected by level, with this formula only taking effect if those don't apply. For example, you could do 10% failure, 80% by formula, and 10% success to keep things from being too sure either way even when levels are very high or low. On the other hand, this flattening makes the benefit of extra advantage levels even less exciting.
Second, we could allow for gradations of success/failure, and treat the coin pools I used to explain that math like dice pools a bit. An in-between could require linearly more success flips to achieve the next higher grade of success at each grade. For example, simple success on a crit role might mean dealing 1.5x damage, but if you succeed on 2 of your flips, you get 9/4 damage, or on 4 flips 27/8, or on 7 flips 81/16. In this world, stacking crit levels might be a viable build, and just giving up on armor would be super dangerous. In the particular case I was using this for just now, I can't easily do gradations of success (that's the reason I turned to probabilities in the first place) but I think I'd favor this approach when feasible.
The main innovation here over simple dice pools is how to handle situations where the number of dice should be negative. I'm almost certain it's not a truly novel innovation though, and some RPG fan can point out which system already does this (please actually do this, I'm an RPG nerd too at heart).
I'll leave this with one more tweak we could do: what if the number 2 in the probability equation were 3, or 2/3? I think this has a similar effect to just scaling all the modifiers a bit, but the algebra escapes me in this moment and I'm a bit lazy. In any case, reducing the base of the probability exponent should let you get a few more gradations near 50%, which is probably a good thing, since the default goes from 25% straight to 50% and then to 75% with no integer stops in between.

Our statement on the rise of #POX (party of xenophobes and fascists), the threat they pose, reasons for their success, strategies, good and bad, for countering them.
Steady State Manchester | For local prosperity, justice & ecological safety

The right wing threat and a viable future
A statement from Steady State Manchester Introduction With horror we see the electoral success of POX, The party of Xenophobes and Fascists (which, without irony, actually calls itself Reform UK1).…

The Army parade was a massive success (Christopher Tremoglie/Washington Examiner)
https://www.washingtonexaminer.com/opinion/beltway-confidential/3444045/army-parade-was-massive-success/
http://www.memeorandum.com/250617/p53#a250617p53

'Can't Take You to a Black Department Store': Pastor Jamal Bryant Says He's 'Embarrassed' By the Success of the Target Boycott; Says Company Leaders Refused to Meet with Him
https://atlantablackstar.com/2025/06/17/pastor-jamal-bryant-says-hes-embarrassed-by-the-success-of-the-target-boycott-says-company-leaders-refused-to-meet/

Impact of a Deployed LLM Survey Creation Tool through the IS Success Model
Peng Jiang, Vinicius Cezar Monteiro de Lira, Antonio Maiorino
https://arxiv.org/abs/2506.14809

Impact of a Deployed LLM Survey Creation Tool through the IS Success Model
Surveys are a cornerstone of Information Systems (IS) research, yet creating high-quality surveys remains labor-intensive, requiring both domain expertise and methodological rigor. With the evolution of large language models (LLMs), new opportunities emerge to automate survey generation. This paper presents the real-world deployment of an LLM-powered system designed to accelerate data collection while maintaining survey quality. Deploying such systems in production introduces real-world complex…

New Raiders LB Germaine Pratt Is No Stranger To Success Vs AFC West Clubs https://raiderramble.com/2025/06/16/new-raiders-lb-germaine-pratt-no-stranger-success-vs-afc-west/

Sixteen years ago today, the #OpenStreetMap #Philippines community 🇵🇭 held the first ever mapping party in the country. And our target was the picturesque city of #Tagaytay.

Theoretical Tensions in RLHF: Reconciling Empirical Success with Inconsistencies in Social Choice Theory
Jiancong Xiao, Zhekun Shi, Kaizhao Liu, Qi Long, Weijie J. Su
https://arxiv.org/abs/2506.12350

Theoretical Tensions in RLHF: Reconciling Empirical Success with Inconsistencies in Social Choice Theory
Despite its empirical success, Reinforcement Learning from Human Feedback (RLHF) has been shown to violate almost all the fundamental axioms in social choice theory -- such as majority consistency, pairwise majority consistency, and Condorcet consistency. This raises a foundational question: why does RLHF perform so well in practice if it fails these seemingly essential properties? In this paper, we resolve this paradox by showing that under mild and empirically plausible assumptions on the pre…

SpaceXplode, am i right? 😉
(I do wish their next flight success 🤞)
#SpaceX

I asked people what they think about working with me, here's what they said!
#Social #Design #Illustration #Portfolio

Dallas Cowboys new addition will be 'major piece' to success, NFL personnel man says https://www.si.com/nfl/cowboys/news/dallas-cowboys-new-addition-will-be-major-piece-to-success-nfl-personnel-man-says…

good news: California Assembly passed a bill to require Caltrans to develop a quick-build program for bike lane and intersection safety projects on state highways! it's #AB891 (Zbur)
https://

Quick-Build Legislation Makes It Through the Assembly - Streetsblog California
Following the success of Quick-Build projects in Santa Monica, their Assemblymember wants to allow them on state highways.

Why AI can't possibly make you more productive; long
#AI and "productivity", some thoughts:
Productivity is a concept that isn't entirely meaningless outside the context of capitalism, but it's a concept that is heavily inflected in a capitalist context. In many uses today it effectively means "how much you can satisfy and/or exceed your boss' expectations." This is not really what it should mean: even in an anarchist utopia, people would care about things like how many shirts they can produce in a week, although in an "I'd like to voluntarily help more people" way rather than an "I need to meet this quota to earn my survival" way. But let's roll with this definition for a second, because it's almost certainly what your boss means when they say "productivity", and understanding that word in a different (even if truer) sense is therefore inherently dangerous.
Accepting "productivity" to mean "satisfying your boss' expectations," I will now claim: the use of generative AI cannot increase your productivity.
Before I dive in, it's imperative to note that the big generative models which most people think of as constituting "AI" today are evil. They are 1: pouring fuel on our burning planet, 2: psychologically strip-mining a class of data laborers who are exploited for their precarity, 3: enclosing, exploiting, and polluting the digital commons, and 4: stealing labor from broad classes of people many of whom are otherwise glad to give that labor away for free provided they get a simple acknowledgement in return. Any of these four "ethical issues" should be enough *alone* to cause everyone to simply not use the technology. These ethical issues are the reason that I do not use generative AI right now, except for in extremely extenuating circumstances. These issues are also convincing for a wide range of people I talk to, from experts to those with no computer science background. So before I launch into a critique of the effectiveness of generative AI, I want to emphasize that such a critique should be entirely unnecessary.
But back to my thesis: generative AI cannot increase your productivity, where "productivity" has been defined as "how much you can satisfy and/or exceed your boss' expectations."
Why? In fact, what the fuck? Every AI booster I've met has claimed the opposite. They've given me personal examples of time saved by using generative AI. Some of them even truly believe this. Sometimes I even believe they saved time without horribly compromising on quality (and often, your boss doesn't care about quality anyways if the lack of quality is hard to measure of doesn't seem likely to impact short-term sales/feedback/revenue). So if generative AI genuinely lets you write more emails in a shorter period of time, or close more tickets, or something else along these lines, how can I say it isn't increasing your ability to meet your boss' expectations?
The problem is simple: your boss' expectations are not a fixed target. Never have been. In virtue of being someone who oversees and pays wages to others under capitalism, your boss' game has always been: pay you less than the worth of your labor, so that they can accumulate profit and this more capital to remain in charge instead of being forced into working for a wage themselves. Sure, there are layers of manservant caught in between who aren't fully in this mode, but they are irrelevant to this analysis. It matters not how much you please your manager if your CEO thinks your work is not worth the wages you are being paid. And using AI actively lowers the value of your work relative to your wages.
Why do I say that? It's actually true in several ways. The most obvious: using generative AI lowers the quality of your work, because the work it produces is shot through with errors, and when your job is reduced to proofreading slop, you are bound to tire a bit, relax your diligence, and let some mistakes through. More than you would have if you are actually doing and taking pride in the work. Examples are innumerable and frequent, from journalists to lawyers to programmers, and we laugh at them "haha how stupid to not check whether the books the AI reviewed for you actually existed!" but on a deeper level if we're honest we know we'd eventually make the same mistake ourselves (bonus game: spot the swipe-typing typos I missed in this post; I'm sure there will be some).
But using generative AI also lowers the value of your work in another much more frightening way: in this era of hype, it demonstrates to your boss that you could be replaced by AI. The more you use it, and no matter how much you can see that your human skills are really necessary to correct its mistakes, the more it appears to your boss that they should hire the AI instead of you. Or perhaps retain 10% of the people in roles like yours to manage the AI doing the other 90% of the work. Paradoxically, the *more* you get done in terms of raw output using generative AI, the more it looks to your boss as if there's an opportunity to get enough work done with even fewer expensive humans. Of course, the decision to fire you and lean more heavily into AI isn't really a good one for long-term profits and success, but the modern boss did not get where they are by considering long-term profits. By using AI, you are merely demonstrating your redundancy, and the more you get done with it, the more redundant you seem.
In fact, there's even a third dimension to this: by using generative AI, you're also providing its purveyors with invaluable training data that allows them to make it better at replacing you. It's generally quite shitty right now, but the more use it gets by competent & clever people, the better it can become at the tasks those specific people use it for. Using the currently-popular algorithm family, there are limits to this; I'm not saying it will eventually transcend the mediocrity it's entwined with. But it can absolutely go from underwhelmingly mediocre to almost-reasonably mediocre with the right training data, and data from prompting sessions is both rarer and more useful than the base datasets it's built on.
For all of these reasons, using generative AI in your job is a mistake that will likely lead to your future unemployment. To reiterate, you should already not be using it because it is evil and causes specific and inexcusable harms, but in case like so many you just don't care about those harms, I've just explained to you why for entirely selfish reasons you should not use it.
If you're in a position where your boss is forcing you to use it, my condolences. I suggest leaning into its failures instead of trying to get the most out of it, and as much as possible, showing your boss very clearly how it wastes your time and makes things slower. Also, point out the dangers of legal liability for its mistakes, and make sure your boss is aware of the degree to which any of your AI-eager coworkers are producing low-quality work that harms organizational goals.
Also, if you've read this far and aren't yet of an anarchist mindset, I encourage you to think about the implications of firing 75% of (at least the white-collar) workforce in order to make more profit while fueling the climate crisis and in most cases also propping up dictatorial figureheads in government. When *either* the AI bubble bursts *or* if the techbros get to live out the beginnings of their worker-replacement fantasies, there are going to be an unimaginable number of economically desperate people living in increasingly expensive times. I'm the kind of optimist who thinks that the resulting social crucible, though perhaps through terrible violence, will lead to deep social changes that effectively unseat from power the ultra-rich that continue to drag us all down this destructive path, and I think its worth some thinking now about what you might want the succeeding stable social configuration to look like so you can advocate towards that during points of malleability.
As others have said more eloquently, generative AI *should* be a technology that makes human lives on average easier, and it would be were it developed & controlled by humanists. The only reason that it's not, is that it's developed and controlled by terrible greedy people who use their unfairly hoarded wealth to immiserate the rest of us in order to maintain their dominance. In the long run, for our very survival, we need to depose them, and I look forward to what the term "generative AI" will mean after that finally happens.

Some statistics about all robotic #LunarLanding attempts so far from 1965 to 2025 compiled from https://en.wikipedia.org/wiki/List_of_missions_to_the_Moon and https://scicomm.xyz/@AkaSci@fosstodon.org/114636519291241321 in which I only count those for which descent to the surface had been initiated, not missions lost at launch or on the way - in a nutshell ~70% of all landings by government agencies went well (essentially the same rate 60 years ago and now!) but only ~30% by private companies. Here goes ...
There have been two separate periods of soft lunar landing attempts of ca. a dozen years each, from 1965 to 1976 and 2013 to 2025 (ongoing) with a huge gap between them.
In the first interval there were 20 attempts with 13 successes (Luna 9, 13, 16, 17, 19, 20, 21 and 24 and Surveyor 1, 3, 5, 6 and 7), one partial success (Luna 23, counting as 50%) and 6 failures (Luna 5, 7, 8, 15 and 18 and Surveyor 4), so the success rate was 13.5/20 = 68 %. All missions were by - the Soviet and U.S. - governments.
In the second interval there were so far 14 attempts with 6 full successes (Chang'e-3, 4, 5 and 6, Vikram 2 and Blue Ghost), three partial successes (SLIM, IM-1 and 2, counting as 75%, 50% and 25%, respectively) and 5 failures (Beresheet, Vikram 1, Hakuto-R 1 and 2 and Luna 25) so the success rate was 7.5 / 14 = 54%.
But looking only at the government missions it was 72%, slighly up from 50 years ago. While for the commercial attempts it was only 29%. In total the success rate was 19 (18 government-run) missions out of 34 (28) attempts or 62% but 69% for governments only. And if you throw in the 6 Apollo landings, the total success rate rises to 68% and the government-only rate goes even up to 75%.

Next-User Retrieval: Enhancing Cold-Start Recommendations via Generative Next-User Modeling
Yu-Ting Lan, Yang Huo, Yi Shen, Xiao Yang, Zuotao Liu
https://arxiv.org/abs/2506.15267 …

Next-User Retrieval: Enhancing Cold-Start Recommendations via Generative Next-User Modeling
The item cold-start problem is critical for online recommendation systems, as the success of this phase determines whether high-quality new items can transition to popular ones, receive essential feedback to inspire creators, and thus lead to the long-term retention of creators. However, modern recommendation systems still struggle to address item cold-start challenges due to the heavy reliance on item and historical interactions, which are non-trivial for cold-start items lacking sufficient ex…

Optimizing Length Compression in Large Reasoning Models
Zhengxiang Cheng, Dongping Chen, Mingyang Fu, Tianyi Zhou
https://arxiv.org/abs/2506.14755 https://…

Optimizing Length Compression in Large Reasoning Models
Large Reasoning Models (LRMs) have achieved remarkable success, yet they often suffer from producing unnecessary and verbose reasoning chains. We identify a core aspect of this issue as "invalid thinking" -- models tend to repeatedly double-check their work after having derived the correct answer. To address this specific inefficiency, we move beyond the general principles of Efficacy and Efficiency to propose two new, fine-grained principles: Brevity, which advocates for eliminating redundancy…

Watching the livestream of what we can all do next after #NoKings protests success on Saturday.
(YouTube link, should work for a replay after the livestream completes too.)
https://www.youtube.com/live/ECEbs-zOfdc

A Variational Framework for Improving Naturalness in Generative Spoken Language Models
Li-Wei Chen, Takuya Higuchi, Zakaria Aldeneh, Ahmed Hussen Abdelaziz, Alexander Rudnicky
https://arxiv.org/abs/2506.14767

A Variational Framework for Improving Naturalness in Generative Spoken Language Models
The success of large language models in text processing has inspired their adaptation to speech modeling. However, since speech is continuous and complex, it is often discretized for autoregressive modeling. Speech tokens derived from self-supervised models (known as semantic tokens) typically focus on the linguistic aspects of speech but neglect prosodic information. As a result, models trained on these tokens can generate speech with reduced naturalness. Existing approaches try to fix this by…

Demonstrating Multi-Suction Item Picking at Scale via Multi-Modal Learning of Pick Success
Che Wang, Jeroen van Baar, Chaitanya Mitash, Shuai Li, Dylan Randle, Weiyao Wang, Sumedh Sontakke, Kostas E. Bekris, Kapil Katyal
https://arxiv.org/abs/2506.10359

Demonstrating Multi-Suction Item Picking at Scale via Multi-Modal Learning of Pick Success
This work demonstrates how autonomously learning aspects of robotic operation from sparsely-labeled, real-world data of deployed, engineered solutions at industrial scale can provide with solutions that achieve improved performance. Specifically, it focuses on multi-suction robot picking and performs a comprehensive study on the application of multi-modal visual encoders for predicting the success of candidate robotic picks. Picking diverse items from unstructured piles is an important and chal…

This https://arxiv.org/abs/2505.12638 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_qbi…

ChromFound: Towards A Universal Foundation Model for Single-Cell Chromatin Accessibility Data
The advent of single-cell Assay for Transposase-Accessible Chromatin using sequencing (scATAC-seq) offers an innovative perspective for deciphering regulatory mechanisms by assembling a vast repository of single-cell chromatin accessibility data. While foundation models have achieved significant success in single-cell transcriptomics, there is currently no foundation model for scATAC-seq that supports zero-shot high-quality cell identification and comprehensive multi-omics analysis simultaneous…

Weak TransNet: A Petrov-Galerkin based neural network method for solving elliptic PDEs
Zhihang Xu, Min Wang, Zhu Wang
https://arxiv.org/abs/2506.14812 http…

Weak TransNet: A Petrov-Galerkin based neural network method for solving elliptic PDEs
While deep learning has achieved remarkable success in solving partial differential equations (PDEs), it still faces significant challenges, particularly when the PDE solutions have low regularity or singularities. To address these issues, we propose the Weak TransNet (WTN) method, based on a Petrov-Galerkin formulation, for solving elliptic PDEs in this work, though its framework may extend to other classes of equations. Specifically, the neural feature space defined by TransNet (Zhang et al.,…

The Pitfalls and Potentials of Adding Gene-invariance to Optimal Mixing
Anton Bouter, Dirk Thierens, Peter A. N. Bosman
https://arxiv.org/abs/2506.15222 ht…

The Pitfalls and Potentials of Adding Gene-invariance to Optimal Mixing
Optimal Mixing (OM) is a variation operator that integrates local search with genetic recombination. EAs with OM are capable of state-of-the-art optimization in discrete spaces, offering significant advantages over classic recombination-based EAs. This success is partly due to high selection pressure that drives rapid convergence. However, this can also negatively impact population diversity, complicating the solving of hierarchical problems, which feature multiple layers of complexity. While t…

Asymptotically Smaller Encodings for Graph Problems and Scheduling
Bernardo Subercaseaux
https://arxiv.org/abs/2506.14042 https://arx…

Asymptotically Smaller Encodings for Graph Problems and Scheduling
We show how several graph problems (e.g., vertex-cover, independent-set, $k$-coloring) can be encoded into CNF using only $O(|V|^2 / \lg |V|)$ many clauses, as opposed to the $Ω(|V|^2)$ constraints used by standard encodings. This somewhat surprising result is a simple consequence of a result of Erdős, Chung, and Spencer (1983) about biclique coverings of graphs, and opens theoretical avenues to understand the success of "Bounded Variable Addition'' (Manthey, Heu…

MLDebugging: Towards Benchmarking Code Debugging Across Multi-Library Scenarios
Jinyang Huang, Xiachong Feng, Qiguang Chen, Hanjie Zhao, Zihui Cheng, Jiesong Bai, Jingxuan Zhou, Min Li, Libo Qin
https://arxiv.org/abs/2506.13824

MLDebugging: Towards Benchmarking Code Debugging Across Multi-Library Scenarios
Code debugging is a crucial task in software engineering, which attracts increasing attention. While remarkable success has been made in the era of large language models (LLMs), current research still focuses on the simple no-library or single-library setting, ignoring the complex multi-library scenario in real-world applications. To address this limitation, we make the first attempt to introduce MLDebugging (Multi-Library Debugging), a comprehensive benchmark designed to assess debugging chall…

We all perform tasks in our day-to-day work that are considered 'non-promotable' – these are crucial for project success, but they won't get you promoted. This is commonly known as 'glue work', a term coined by Tanya Reilly. Join Fatima Taj at this year's Berlin Buzzwords as she shares her personal experience of narrowly avoiding the trap of being permanently stuck with glue work, and explains how to handle similar situations.
Learn more:

ViSTA: Visual Storytelling using Multi-modal Adapters for Text-to-Image Diffusion Models
Sibo Dong, Ismail Shaheen, Maggie Shen, Rupayan Mallick, Sarah Adel Bargal
https://arxiv.org/abs/2506.12198

ViSTA: Visual Storytelling using Multi-modal Adapters for Text-to-Image Diffusion Models
Text-to-image diffusion models have achieved remarkable success, yet generating coherent image sequences for visual storytelling remains challenging. A key challenge is effectively leveraging all previous text-image pairs, referred to as history text-image pairs, which provide contextual information for maintaining consistency across frames. Existing auto-regressive methods condition on all past image-text pairs but require extensive training, while training-free subject-specific approaches ens…

Stephen Miller is becoming a victim of his own success (Hayes Brown/MSNBC)
https://www.msnbc.com/opinion/msnbc-opinion/stephen-miller-ice-deportation-rcna213491
http://www.memeorandum.com/250618/p26#a250618p26

BraTS orchestrator : Democratizing and Disseminating state-of-the-art brain tumor image analysis
Florian Kofler, Marcel Rosier, Mehdi Astaraki, Ujjwal Baid, Hendrik M\"oller, Josef A. Buchner, Felix Steinbauer, Eva Oswald, Ezequiel de la Rosa, Ivan Ezhov, Constantin von See, Jan Kirschke, Anton Schmick, Sarthak Pati, Akis Linardos, Carla Pitarch, Sanyukta Adap, Jeffrey Rudie, Maria Correia de Verdier, Rachit Saluja, Evan Calabrese, Dominic LaBella, Mariam Aboian, Ahmed W. Moawad, …

BraTS orchestrator : Democratizing and Disseminating state-of-the-art brain tumor image analysis
The Brain Tumor Segmentation (BraTS) cluster of challenges has significantly advanced brain tumor image analysis by providing large, curated datasets and addressing clinically relevant tasks. However, despite its success and popularity, algorithms and models developed through BraTS have seen limited adoption in both scientific and clinical communities. To accelerate their dissemination, we introduce BraTS orchestrator, an open-source Python package that provides seamless access to state-of-the-…

Why not download an app and go through a digital identity verification process to... have quick access to your National Insurance number?
https://www.gov.uk/government/news/be-summer-job-ready-with-the-hmrc-app

Be summer-job ready with the HMRC app
Download the HMRC app for summer job success

nice quote from Vivek Murthy
(interviewed by Eric Topol)
"... a triad of success ... a triad of fulfillment.
"... how does society define success for you? And they would tell me some version of money, fame, and power. If you had all three of those, then you really hit the lottery and people will write books about you, make documentaries about you, you'll have made it. But ... The triad to fulfillment is rooted in relationships, service, and purpose"
#life #success

Things to watch every 6 months
#mastodonadmin

Understanding multi-fidelity training of machine-learned force-fields
John L. A. Gardner, Hannes Schulz, Jean Helie, Lixin Sun, Gregor N. C. Simm
https://arxiv.org/abs/2506.14963 …

Understanding multi-fidelity training of machine-learned force-fields
Effectively leveraging data from multiple quantum-chemical methods is essential for building machine-learned force fields (MLFFs) that are applicable to a wide range of chemical systems. This study systematically investigates two multi-fidelity training strategies, pre-training/fine-tuning and multi-headed training, to elucidate the mechanisms underpinning their success. We identify key factors driving the efficacy of pre-training followed by fine-tuning, but find that internal representations …

An Atomic Cluster Expansion Potential for Twisted Multilayer Graphene
Yangshuai Wang, Drake Clark, Sambit Das, Ziyan Zhu, Daniel Massatt, Vikram Gavini, Mitchell Luskin, Christoph Ortner
https://arxiv.org/abs/2506.15061

An Atomic Cluster Expansion Potential for Twisted Multilayer Graphene
Twisted multilayer graphene, characterized by its moire patterns arising from inter-layer rotational misalignment, serves as a rich platform for exploring quantum phenomena. While first-principles calculations are computationally prohibitive and empirical interatomic potentials often lack accuracy, machine-learning interatomic potentials (MLIPs) present a promising alternative, offering (near-)DFT accuracy at a significantly reduced computational cost. Despite their success in two-dimensional m…

The Future of Solar modelling: requirements for a new generation of solar models
Gael Buldgen, Gloria. Canocchi, Arthur. Le Saux, Vladimir A. Baturin, Regner Trampedach, Anna V. Oreshina, Sergey V. Ayukov, Anil Pradhan, Jean-Christophe Pain, Masanobu Kunitomo, Thierry Appourchaux, Rafael A. Garcia, Morgan Deal, Nicolas Grevesse, Arlette Noels, Joergen Christensen-Dalsgaard, Tristan Guillot, Devesh Nandal, J\'er\^ome B\'etrisey, Christophe Blancard, James Colgan, Philippe Coss\&…

The Future of Solar modelling: requirements for a new generation of solar models
Helioseismology and solar modelling have enjoyed a golden era thanks to decades-long surveys from ground-based networks such as for example GONG, BiSON, IRIS and the SOHO and SDO space missions which have provided high-quality helioseismic observations that supplemented photometric, gravitational, size and shape, limb-darkening and spectroscopic constraints as well as measurements of neutrino fluxes. However, the success of solar models is also deeply rooted in progress in fundamental physics (…

Trump is a bit like the mosquito who was out flying for the first time and when he got back his parents asked him: "So, how did it go?" And he replied: "It was an amazing success! Everybody applauded!"
#trump #humor

InverTune: Removing Backdoors from Multimodal Contrastive Learning Models via Trigger Inversion and Activation Tuning
Mengyuan Sun, Yu Li, Yuchen Liu, Bo Du, Yunjie Ge
https://arxiv.org/abs/2506.12411

InverTune: Removing Backdoors from Multimodal Contrastive Learning Models via Trigger Inversion and Activation Tuning
Multimodal contrastive learning models like CLIP have demonstrated remarkable vision-language alignment capabilities, yet their vulnerability to backdoor attacks poses critical security risks. Attackers can implant latent triggers that persist through downstream tasks, enabling malicious control of model behavior upon trigger presentation. Despite great success in recent defense mechanisms, they remain impractical due to strong assumptions about attacker knowledge or excessive clean data requir…

So the basic idea is that we first compute a "level" for whatever interaction, by adding beneficial modifiers and subtracting harmful ones. Imagine most modifiers are smallish integers like 2 or -3 (though they can be non-integers too). Each level can be thought of as making things twice as good/bad, although this only applies directly when they're balanced. The actual formula starts with a 50/50 chance of "success" at level 0, and then each positive level halves the chance of failure, or if the levels are negative, each negative level halves the chance of success (note that halving the chance of failure is not the same as doubling the chance of success).
The intuitive explanation is that you start with a coin flip. Then if the level is positive, you flip that many additional coins and succeed if any single coin succeeds, but it the level is negative, you have to flip that many additional coins and succeed only if *all* flips succeed.
For example, if I have a dagger with 5 crit chance, and I attack an opponent with no armor modifiers, I'd have to win any 1 of 6 coin flips to score a crit (p = 1 - (1/(2^6)) = 63/64. Increasing my crit modifier by 1 ups my chances only slightly, to 127/128. This is obviously pretty poor return, indicating that the 5 I already have is very strong. If the opponent had armor with -3 to crits, the interaction is now level 2, so the crit chance is 7/8, which is still pretty good. We can see from these examples that the basic system
rewards a small level advantage a lot, but the rewards diminish rapidly. The system has a few avenues for tweaking how it works though, that can let us modify this. There's also a potential benefit (though sometimes drawback) that no matter what the level gap, there's an effective limit to how much the interaction swings.

Belief propagation for networks with loops: The neighborhoods-intersections-based method
Pedro Hack
https://arxiv.org/abs/2506.13791 https://

Belief propagation for networks with loops: The neighborhoods-intersections-based method
In order to diminish the damaging effect of loops on belief propagation (BP), the first explicit version of generalized BP for networks, the KCN-method, was recently introduced. Despite its success, the KCN-method spends computational resources inefficiently. Such inefficiencies can quickly turn the exact application of the method unfeasible, since its time complexity increases exponentially with them. This affects for instance tree networks, for which, despite not offering any accuracy advanta…

Nick Chubb to Texans: Can former All-Pro rusher continue success of RBs going to new teams post-ACL injury?

https://www.cbssports.com/nfl/news/n…

Towards Employing FPGA and ASIP Acceleration to Enable Onboard AI/ML in Space Applications
Vasileios Leon, George Lentaris, Dimitrios Soudris, Simon Vellas, Mathieu Bernou
https://arxiv.org/abs/2506.12970

Towards Employing FPGA and ASIP Acceleration to Enable Onboard AI/ML in Space Applications
The success of AI/ML in terrestrial applications and the commercialization of space are now paving the way for the advent of AI/ML in satellites. However, the limited processing power of classical onboard processors drives the community towards extending the use of FPGAs in space with both rad-hard and Commercial-Off-The-Shelf devices. The increased performance of FPGAs can be complemented with VPU or TPU ASIP co-processors to further facilitate high-level AI development and in-flight reconfigu…

My aging #Ubiquiti Edgerouter-X is failing. So, I spent 3 hours yesterday trying to set up a new #Unifi Gateway Max without success. I'm returning it.
I decided to just get a #miniPC with dua…

ChromFound: Towards A Universal Foundation Model for Single-Cell Chromatin Accessibility Data
Yifeng Jiao, Yuchen Liu, Yu Zhang, Xin Guo, Yushuai Wu, Chen Jiang, Jiyang Li, Hongwei Zhang, Limei Han, Xin Gao, Yuan Qi, Yuan Cheng
https://arxiv.org/abs/2505.12638

ChromFound: Towards A Universal Foundation Model for Single-Cell Chromatin Accessibility Data
The advent of single-cell Assay for Transposase-Accessible Chromatin using sequencing (scATAC-seq) offers an innovative perspective for deciphering regulatory mechanisms by assembling a vast repository of single-cell chromatin accessibility data. While foundation models have achieved significant success in single-cell transcriptomics, there is currently no foundation model for scATAC-seq that supports zero-shot high-quality cell identification and comprehensive multi-omics analysis simultaneous…

Med-REFL: Medical Reasoning Enhancement via Self-Corrected Fine-grained Reflection
Zongxian Yang, Jiayu Qian, Zegao Peng, Haoyu Zhang, Zhi-An Huang
https://arxiv.org/abs/2506.13793

Med-REFL: Medical Reasoning Enhancement via Self-Corrected Fine-grained Reflection
Large reasoning models have recently made significant strides in mathematical and code reasoning, yet their success has not transferred smoothly to the medical domain. While multiple factors contribute to this disparity, a critical issue is the inadequate focus on the quality of intermediate reflection steps, which is particularly crucial in high-stakes medical scenarios. To address this challenge, we propose Med-REFL, a \underline{\textbf{Med}}ical \underline{\textbf{R}}easoning \underline{\te…

Wahey! Success on my Javastation Krups, this is a Debian Sarge debootstrapped userspace running over NFS.
#retrocomputing

Nexus of Team Collaboration Stability on Mega Construction Project Success in Electric Vehicle Manufacturing Enterprises: The Moderating Role of Human-AI Integration
Jun Cui
https://arxiv.org/abs/2506.06375

Nexus of Team Collaboration Stability on Mega Construction Project Success in Electric Vehicle Manufacturing Enterprises: The Moderating Role of Human-AI Integration
This study investigates how team collaboration stability influences the success of mega construction projects in electric vehicle manufacturing enterprises, with human-AI integration as a moderating variable. Using structural equation modeling (SEM) with data from 187 project teams across China's electric vehicle sector, results indicate that team collaboration stability significantly enhances project success. The moderating effect of human-AI integration strengthens this relationship, suggesti…

Balkonkraftwerk is the German success story of plug-in solar energy (one or two solar panels, inverter with 🔌), which can be self-installed by apartment dwellers and has democratised solar energy. There are now officially over one million of these installations in Germany.
(cartoon ©️ Til Mette)
https://www.…

3 sleepers critical to Cowboys success this season https://insidethestar.com/3-sleepers-critical-to-cowboys-success-this-season

The right wing threat and a viable future
A statement from Steady State Manchester Introduction With horror we see the electoral success of POX, The party of Xenophobes and Fascists (which, without irony, actually calls itself Reform UK1). In the May election round they gained 677 more councillors, took control of 8 councils, landed a regional Mayor and gained another MP, in a formerly safe Labour seat.

The right wing threat and a viable future
A statement from Steady State Manchester Introduction With horror we see the electoral success of POX, The party of Xenophobes and Fascists (which, without irony, actually calls itself Reform UK1).…

I don’t want to dwell •too• much on the shadenfreude of Trump’s pitiful parade, but I do think it’s worth just taking this in:
Trump’s political success has always hinged on him knowing how to be a reality TV star, knowing better than the news orgs themselves that political “news” is in fact reality TV, knowing how to manipulate that reality TV show.
And today, he failed at that completely. His reality show sucked. The news of the day just ran right over him. https://kolektiva.social/@Voline/114685755430626127

#generativeAI experiment : for the page of software package, I am trying to generate a woman computer vision scientist in a superman costume...
I tried different ways to force the generated images to depict a woman, but with no success : I only get the same male scientist!

Know What You Don't Know: Uncertainty Calibration of Process Reward Models
Young-Jin Park, Kristjan Greenewald, Kaveh Alim, Hao Wang, Navid Azizan
https://arxiv.org/abs/2506.09338

Know What You Don't Know: Uncertainty Calibration of Process Reward Models
Process reward models (PRMs) play a central role in guiding inference-time scaling algorithms for large language models (LLMs). However, we observe that even state-of-the-art PRMs can be poorly calibrated and often overestimate success probabilities. To address this, we present a calibration approach, performed via quantile regression, that adjusts PRM outputs to better align with true success probabilities. Leveraging these calibrated success estimates and their associated confidence bounds, w…

A weaponized AI chatbot is flooding Canadian city councils with climate misinformation, and already having some 'success'.
https://www.nationalobserver.com/2025/05/28/investigations/weaponized-ai-chatbot-…

A weaponized AI chatbot is flooding city councils with climate misinformation
Internal instructions accessed by Canada’s National Observer show that the chatbot produces tailored scripts, petitions, reports and even speeches for council chambers. The messaging is all framed to resonate with municipal officials’ duty to represent local interests. The chatbot drops the cost of misinformation to "close to zero."

Limiting distributions of ratios of Binomial random variables
Adriel Barretto, Zachary Lubberts
https://arxiv.org/abs/2506.13071 https://

Limiting distributions of ratios of Binomial random variables
We consider the limiting distribution of the quantity $X^s/(X+Y)^r$, where $X$ and $Y$ are two independent Binomial random variables with a common success probability and a number of trials $n$ and $m$, respectively, and $r,s$ are positive real numbers. Under several settings, we prove that this converges to a Normal distribution with a given mean and variance, and demonstrate these theoretical results through simulations.

What Drives Team Success? Large-Scale Evidence on the Role of the Team Player Effect
Nico Elbert, Alicia von Schenk, Fabian Kosse, Victor Klockmann, Nikolai Stein, Christoph Flath
https://arxiv.org/abs/2506.04475

What Drives Team Success? Large-Scale Evidence on the Role of the Team Player Effect
Effective teamwork is essential in structured, performance-driven environments, from professional organizations to high-stakes competitive settings. As tasks grow more complex, achieving high performance requires not only technical proficiency but also strong interpersonal skills that allow individuals to coordinate effectively within teams. While prior research has identified social skills and familiarity as key drivers of team success, their joint effects -- particularly in temporary teams --…

YouTube releases its 2024 US Impact Report, which says YouTube's creative ecosystem contributed $55B to US GDP in 2024 and supported 490K full-time jobs (Alexandra Veitch/YouTube Official Blog)
https://blog.youtube/news-and-events/2024-us-youtube-impact-re…

2024 U.S. YouTube Impact Report
Learn how YouTube empowers creators and fuels the US economy. Our new report highlights a $55B US GDP contribution and 490K jobs supported, showcasing two decades of shared growth and creator success.

How oxygen influences the catalytic activity of iron during carbon nanotube nucleation
Ben McLean, Alister J. Page, Feng Ding
https://arxiv.org/abs/2506.11632

How oxygen influences the catalytic activity of iron during carbon nanotube nucleation
The catalytic activity of metal nanoparticles toward nucleation of single-walled carbon nanotubes (SWCNTs) is fundamental to achieving structure-controlled growth using catalytic chemical vapor deposition (CVD). Despite the success of oxidized catalysts in SWCNT growth, there is a lack of understanding regarding how oxygen influences the catalysts and the nucleation process. Quantum chemical molecular dynamics (MD) simulations employing density functional tight binding (DFTB) demonstrate that t…

OH – food for thought: “They've found ways to be successful critics – emphasis on successful. They get invited to conferences, win awards, maintain influence … that success might come from never truly threatening the system.
Maybe aggressive rhetoric and a bridge-burning approach isn’t a bug but a feature. When the house is on fire, you don’t negotiate with the arsonists about fire safety regulations. You sound the alarm, loudly and repeatedly.
In a truly existential fight, w…

This is what midi hell looks like... but I'm having success getting it sorted 🥵

Kelce 'only interested' in SB rings, not AFC bling https://www.espn.com/nfl/story/_/id/45539848/chiefs-travis-kelce-only-interested-super-bowl-rings

Trump is a bit like the mosquito who was out flying for the first time and when he got back his parents asked him: "So, how did it go?" And he replied: "It was an amazing success! Everybody applauded!"
#trump #humor

What success looks like... "could not be found". This is the Austrian railway's WiFi login. I can use the Internet now though. :D

Trash panda update: Didn't set any more boobytraps. Sprinkled cayenne pepper powder along a bunch of their travel routes (but not fully ringing the yard) which didn't seem to do a whole lot. Maybe they just walked around it, or I need to use more? I'll try doing the entire fence line tonight and see if it's more effective if I make sure not to leave any gaps.
I tried the nerf gun for what I think will probably be the last time. Shot at one of them and the cheeky bugger …

Alphabet Index Mapping: Jailbreaking LLMs through Semantic Dissimilarity
Bilal Saleh Husain
https://arxiv.org/abs/2506.12685 https://…

Alphabet Index Mapping: Jailbreaking LLMs through Semantic Dissimilarity
Large Language Models (LLMs) have demonstrated remarkable capabilities, yet their susceptibility to adversarial attacks, particularly jailbreaking, poses significant safety and ethical concerns. While numerous jailbreak methods exist, many suffer from computational expense, high token usage, or complex decoding schemes. Liu et al. (2024) introduced FlipAttack, a black-box method that achieves high attack success rates (ASR) through simple prompt manipulation. This paper investigates the underly…

Our new statement on the rise of #POX (party of xenophobes and fascists), the threat they pose, reasons for their success, strategies, good and bad, for countering them.
Steady State Manchester | For local prosperity, justice & ecological safety

The right wing threat and a viable future
A statement from Steady State Manchester Introduction With horror we see the electoral success of POX, The party of Xenophobes and Fascists (which, without irony, actually calls itself Reform UK1).…

Of *course* they did. And it's beyond ironic that the SBC has someone labeled an "ethicist" who's focused on this instead of fixing the sex abuse problem the conference has. #Obergefell

Southern Baptists Call to Overturn Supreme Court Ruling on Same-Sex Marriage
The nation’s largest Protestant denomination was motivated by conservative Christians’ success in reversing Roe v. Wade.

This is really great and will only increase the success of this AI tool. I can recommend NotebookLM, try it.
https://www.theverge.com/news/678915/google-notebooklm-share-public-link

Google’s NotebookLM now lets you share your notebook — and AI podcasts — publicly
Google’s AI-powered note-taking app, NotebookLM, will now let others view your notebooks with public links.

This https://arxiv.org/abs/2505.01997 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csLG_…

Restoring Calibration for Aligned Large Language Models: A Calibration-Aware Fine-Tuning Approach
One of the key technologies for the success of Large Language Models (LLMs) is preference alignment. However, a notable side effect of preference alignment is poor calibration: while the pre-trained models are typically well-calibrated, LLMs tend to become poorly calibrated after alignment with human preferences. In this paper, we investigate why preference alignment affects calibration and how to address this issue. For the first question, we observe that the preference collapse issue in align…

Las Vegas Raiders celebrate success of 5th annual INSPIRE event held at Intermountain Health Performance Center, presented by Allegiant https://www.raiders.com/news/las-vegas-rai

Ising versus infinite randomness criticality in arrays of Rydberg atoms trapped with non-perfect tweezers
Jose Soto-Garcia, Natalia Chepiga
https://arxiv.org/abs/2506.11985

Ising versus infinite randomness criticality in arrays of Rydberg atoms trapped with non-perfect tweezers
Chains of Rydberg atoms have emerged as an amazing platform for simulating quantum physics in low dimensions. This remarkable success is due to the versatility of lattice geometries achieved by trapping neutral atoms with optical tweezers. On a given lattice, the competition between the repulsive van der Waals potential and the detuning of the laser frequency brings the atoms to highly excited Rydberg states, leading to a variety of exotic phases and quantum phase transitions. Experiments on th…

Learning to Optimize Package Picking for Large-Scale, Real-World Robot Induction
Shuai Li, Azarakhsh Keipour, Sicong Zhao, Srinath Rajagopalan, Charles Swan, Kostas E. Bekris
https://arxiv.org/abs/2506.09765

Learning to Optimize Package Picking for Large-Scale, Real-World Robot Induction
Warehouse automation plays a pivotal role in enhancing operational efficiency, minimizing costs, and improving resilience to workforce variability. While prior research has demonstrated the potential of machine learning (ML) models to increase picking success rates in large-scale robotic fleets by prioritizing high-probability picks and packages, these efforts primarily focused on predicting success probabilities for picks sampled using heuristic methods. Limited attention has been given, howev…

The secret to Dak Prescott’s success in 2025 https://insidethestar.com/the-secret-to-dak-prescotts-success-in-2025

Exploiting the Exact Denoising Posterior Score in Training-Free Guidance of Diffusion Models
Gregory Bellchambers
https://arxiv.org/abs/2506.13614 https://…

Exploiting the Exact Denoising Posterior Score in Training-Free Guidance of Diffusion Models
The success of diffusion models has driven interest in performing conditional sampling via training-free guidance of the denoising process to solve image restoration and other inverse problems. A popular class of methods, based on Diffusion Posterior Sampling (DPS), attempts to approximate the intractable posterior score function directly. In this work, we present a novel expression for the exact posterior score for purely denoising tasks that is tractable in terms of the unconditional score fu…

Russell Wilson says Giants drafting Jaxson Dart 'doesn't change anything at all,' veteran focused on 'success'

https://www.cb…

Las Vegas Raiders celebrate success of 5th annual INSPIRE event held at Intermountain Health Performance Center, presented by Allegiant https://www.raiders.com/news/las-vegas-rai

A look at the Salt Lake Tribune under CEO and top editor Lauren Gustus, who is spearheading a campaign to raise $1M in 2025 to end its paywall (Rick Edmonds/Poynter)
https://www.poynter.org/business-work/2025/salt-lake-tribune-paywal…

At the nonprofit Salt Lake Tribune, a turnaround and now a big gamble - Poynter
The Tribune’s success has made it a poster child for going nonprofit. Will its attempts to shake off its paywall find similar success?

Deep Symmetric Autoencoders from the Eckart-Young-Schmidt Perspective
Simone Brivio, Nicola Rares Franco
https://arxiv.org/abs/2506.11641 https://

Deep Symmetric Autoencoders from the Eckart-Young-Schmidt Perspective
Deep autoencoders have become a fundamental tool in various machine learning applications, ranging from dimensionality reduction and reduced order modeling of partial differential equations to anomaly detection and neural machine translation. Despite their empirical success, a solid theoretical foundation for their expressiveness remains elusive, particularly when compared to classical projection-based techniques. In this work, we aim to take a step forward in this direction by presenting a com…

Trump is frustrated by his own success on immigration (Eric Levitz/Vox)
https://www.vox.com/politics/416526/la-protests-trump-ice-deportations-border-crossings
http://www.memeorandum.com/250612/p15#a250612p15

Text-Aware Image Restoration with Diffusion Models
Jaewon Min, Jin Hyeon Kim, Paul Hyunbin Cho, Jaeeun Lee, Jihye Park, Minkyu Park, Sangpil Kim, Hyunhee Park, Seungryong Kim
https://arxiv.org/abs/2506.09993

Text-Aware Image Restoration with Diffusion Models
Image restoration aims to recover degraded images. However, existing diffusion-based restoration methods, despite great success in natural image restoration, often struggle to faithfully reconstruct textual regions in degraded images. Those methods frequently generate plausible but incorrect text-like patterns, a phenomenon we refer to as text-image hallucination. In this paper, we introduce Text-Aware Image Restoration (TAIR), a novel restoration task that requires the simultaneous recovery of…

Variational Learning Finds Flatter Solutions at the Edge of Stability
Avrajit Ghosh, Bai Cong, Rio Yokota, Saiprasad Ravishankar, Rongrong Wang, Molei Tao, Mohammad Emtiyaz Khan, Thomas M\"ollenhoff
https://arxiv.org/abs/2506.12903

Variational Learning Finds Flatter Solutions at the Edge of Stability
Variational Learning (VL) has recently gained popularity for training deep neural networks and is competitive to standard learning methods. Part of its empirical success can be explained by theories such as PAC-Bayes bounds, minimum description length and marginal likelihood, but there are few tools to unravel the implicit regularization in play. Here, we analyze the implicit regularization of VL through the Edge of Stability (EoS) framework. EoS has previously been used to show that gradient d…

How an up-and-coming NFL agent (in Minneapolis!) finds his footing in a cut-throat league https://www.nytimes.com/athletic/6415008/2025/06/17/nfl-agent-minnesota-vikings-minneapolis-jonathan-allen/

This https://arxiv.org/abs/2309.01754 has been replaced.
link: https://scholar.google.com/scholar?q=a

On the success probability of the quantum algorithm for the short DLP
Ekerå and Håstad have introduced a variation of Shor's algorithm for the discrete logarithm problem (DLP). Unlike Shor's original algorithm, Ekerå-Håstad's algorithm solves the short DLP in groups of unknown order. In this work, we prove a lower bound on the probability of Ekerå-Håstad's algorithm recovering the short logarithm $d$ in a single run. By our bound, the success probability can easily be pushed as high as $1 - 10^{-10}$ for any short $d$. A key to achieving such a high success…

This https://arxiv.org/abs/2505.18344 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csLG_…

Sample Complexity of Diffusion Model Training Without Empirical Risk Minimizer Access
Diffusion models have demonstrated state-of-the-art performance across vision, language, and scientific domains. Despite their empirical success, prior theoretical analyses of the sample complexity suffer from poor scaling with input data dimension or rely on unrealistic assumptions such as access to exact empirical risk minimizers. In this work, we provide a principled analysis of score estimation, establishing a sample complexity bound of $\widetilde{\mathcal{O}}(ε^{-6})$. Our approach lever…

Patriots All-Quarter Century Team: Tom Brady, Bill Belichick and Rob Gronkowski headline dynastic run

https://www.cbssports.com/nfl/news/patriot

SAFE: Multitask Failure Detection for Vision-Language-Action Models
Qiao Gu, Yuanliang Ju, Shengxiang Sun, Igor Gilitschenski, Haruki Nishimura, Masha Itkina, Florian Shkurti
https://arxiv.org/abs/2506.09937

SAFE: Multitask Failure Detection for Vision-Language-Action Models
While vision-language-action models (VLAs) have shown promising robotic behaviors across a diverse set of manipulation tasks, they achieve limited success rates when deployed on novel tasks out-of-the-box. To allow these policies to safely interact with their environments, we need a failure detector that gives a timely alert such that the robot can stop, backtrack, or ask for help. However, existing failure detectors are trained and tested only on one or a few specific tasks, while VLAs require…

3 most important Dallas Cowboys on defense not named Micah Parsons https://www.si.com/nfl/cowboys/news/3-most-important-dallas-cowboys-on-defense-not-named-micah-parsons

SOFT: Selective Data Obfuscation for Protecting LLM Fine-tuning against Membership Inference Attacks
Kaiyuan Zhang, Siyuan Cheng, Hanxi Guo, Yuetian Chen, Zian Su, Shengwei An, Yuntao Du, Charles Fleming, Ashish Kundu, Xiangyu Zhang, Ninghui Li
https://arxiv.org/abs/2506.10424

SOFT: Selective Data Obfuscation for Protecting LLM Fine-tuning against Membership Inference Attacks
Large language models (LLMs) have achieved remarkable success and are widely adopted for diverse applications. However, fine-tuning these models often involves private or sensitive information, raising critical privacy concerns. In this work, we conduct the first comprehensive study evaluating the vulnerability of fine-tuned LLMs to membership inference attacks (MIAs). Our empirical analysis demonstrates that MIAs exploit the loss reduction during fine-tuning, making them highly effective in re…

Bills WR Keon Coleman gives harsh self-assessment of rookie season: 'When that (expletive) trash, you got to be better' https://www.nfl.com/news/bills-wr-keon-coleman-gives-harsh-self-assessment-of-rookie-season-yo…

4 most important Dallas Cowboys on offense not named Dak Prescott or Ceedee Lamb https://www.si.com/nfl/cowboys/news/4-most-important-dallas-cowboys-on-offense-not-named-dak-prescott-or-ceedee-lamb

Logarithmic Smoothing for Adaptive PAC-Bayesian Off-Policy Learning
Maxime Haddouche, Otmane Sakhi
https://arxiv.org/abs/2506.10664 https://

Logarithmic Smoothing for Adaptive PAC-Bayesian Off-Policy Learning
Off-policy learning serves as the primary framework for learning optimal policies from logged interactions collected under a static behavior policy. In this work, we investigate the more practical and flexible setting of adaptive off-policy learning, where policies are iteratively refined and re-deployed to collect higher-quality data. Building on the success of PAC-Bayesian learning with Logarithmic Smoothing (LS) in static settings, we extend this framework to the adaptive scenario using tool…

TooBadRL: Trigger Optimization to Boost Effectiveness of Backdoor Attacks on Deep Reinforcement Learning
Songze Li, Mingxuan Zhang, Oubo Ma, Kang Wei, Shouling Ji
https://arxiv.org/abs/2506.09562

TooBadRL: Trigger Optimization to Boost Effectiveness of Backdoor Attacks on Deep Reinforcement Learning
Deep reinforcement learning (DRL) has achieved remarkable success in a wide range of sequential decision-making domains, including robotics, healthcare, smart grids, and finance. Recent research demonstrates that attackers can efficiently exploit system vulnerabilities during the training phase to execute backdoor attacks, producing malicious actions when specific trigger patterns are present in the state observations. However, most existing backdoor attacks rely primarily on simplistic and heu…

Cowboys' Dak Prescott labeled NFL's 'worst' in unfortunate category https://www.si.com/nfl/cowboys/news/dallas-cowboys-dak-prescott-labeled-nfl-worst-unfortunate-category

Mailbag: Position that needs to improve most? https://www.dallascowboys.com/news/mailbag-position-that-needs-to-improve-most

16 players to trade for or trade away in dynasty leagues https://www.espn.com/fantasy/football/story/_/id/45478615/2025-fantasy-football-trade-targets-dynasty

Mailbag: Position that needs to improve most? https://www.dallascowboys.com/news/mailbag-position-that-needs-to-improve-most

Dak Prescott makes goal clear about Dallas Cowboys, NFL legacy https://www.si.com/nfl/cowboys/news/dak-prescott-makes-goal-clear-about-dallas-cowboys-nfl-legacy