Tootfinder

Ukraine war latest: Ukrainian retreat looms without US support, ATACMS is ‘the answer,’ says Zelensky: https://benborges.xyz/2024/03/31/ukraine-war-latest.html

Copilot ist böse. Ich hab ihm gesagt, es solle keine Emojis verwenden, und was tut es?
#copilotisevil

Keine Emojis
Here's an answer I got using Microsoft Copilot, the world's first AI-powered answer engine. Select to see the full answer or try it yourself.

House Republicans won’t take ‘no evidence’ for an answer in impeachment inquiry
“To quote Congressman Issa, who was there, it was a big nothing,” said Rep. Jamie Raskin (D-Md.), the top Democrat on the committee.
“Although I would amend it by saying it was a little nothing.
Everything that we learned was what we knew before, which was that President Biden is not involved and has never been involved in his son’s businesses.”

Much of the hearing, he said, involved “pi…

Analysis | House Republicans won’t take ‘no evidence’ for an answer in impeachment inquiry
The closed-door deposition of Hunter Biden involved a lot of denials from someone Republicans refuse to believe.

#potd "My" turn-right-on-red (if you're on a #bicycle) sign was installed! I suggested the location last year, and got a very quick answer from MA46 that they will install it in spring.

Let LLMs Take on the Latest Challenges! A Chinese Dynamic Question Answering Benchmark
Zhikun Xu, Yinghui Li, Ruixue Ding, Xinyu Wang, Boli Chen, Yong Jiang, Xiaodong Deng, Jianxin Ma, Hai-Tao Zheng, Wenlian Lu, Pengjun Xie, Chang Zhou, Fei Huang
https://arxiv.org/abs/2402.19248

Let LLMs Take on the Latest Challenges! A Chinese Dynamic Question Answering Benchmark
How to better evaluate the capabilities of Large Language Models (LLMs) is the focal point and hot topic in current LLMs research. Previous work has noted that due to the extremely high cost of iterative updates of LLMs, they are often unable to answer the latest dynamic questions well. To promote the improvement of Chinese LLMs' ability to answer dynamic questions, in this paper, we introduce CDQA, a Chinese Dynamic QA benchmark containing question-answer pairs related to the latest news on th…

Bumble redesigns its app and adds Opening Moves, to let women ask a question men can answer, a major shift that ends the requirement that women message first (Valeriya Safronova/New York Times)
https://www.nytimes.com/2024/04/30/style/bumble-dating-apps.html

Bumble Tells Women They No Longer Have to Make the First Move
As dissatisfaction with online dating grows, the app that put women in control is shifting course.

Surprised to see the best answer on this is from Danny Sauter. None of them explicitly urge a No vote (boo), but Sauter, Sharon Lai, and Moe Jamil say kind of critical things about it, with Sauter being the most strongly critical.
https://missionlocal.org/2024/02/meet-

Meet the District 3 candidates: ‘Do you support Proposition F? Why or why not?’
Up until the November election, Mission Local will ask each candidate one question per week, and candidates will get 100 words to respond.

I was looking for an answer to the simple question, how large is the trained GPT4 model in Gigabyte, terribly hard to find some estimate ? 🤔
#AI #GPT4 #size

@… not exactly an answer to your question, but it's available on Amazons german site, so at least it isn't US-exclusive
https://amzn.eu/d/4GO68fQ

Note to self: If you are a journalist and want to ask a company one critical, important question, and maybe a few others, it's probably a better idea to skip the few others, or ask them in a later mail. Avoids that they can play games and only answer the lesser important questions and just skip the relevant one.

@… for some reason quote posts don’t seem to work for me anymore… long pressing the answer icon?
Am I holding it wrong?

Having done interviews myself, I know live media tend to call on those who answer requests quickly. This leads to commentators ready to talk about anything anytime. But sometimes, maybe call on a scholar who's dedicated years of their life to understanding a subject instead of the same old political hack? Hearing Tory operatives line up behind the mic on CBC/Radio-Canada to talk Mulroney makes me pine for what historians like Dimitry Anastakis or Penny Bryden could contribute.

Unsolvable Problem Detection: Evaluating Trustworthiness of Vision Language Models
Atsuyuki Miyai, Jingkang Yang, Jingyang Zhang, Yifei Ming, Qing Yu, Go Irie, Yixuan Li, Hai Li, Ziwei Liu, Kiyoharu Aizawa
https://arxiv.org/abs/2403.20331

Unsolvable Problem Detection: Evaluating Trustworthiness of Vision Language Models
This paper introduces a novel and significant challenge for Vision Language Models (VLMs), termed Unsolvable Problem Detection (UPD). UPD examines the VLM's ability to withhold answers when faced with unsolvable problems in the context of Visual Question Answering (VQA) tasks. UPD encompasses three distinct settings: Absent Answer Detection (AAD), Incompatible Answer Set Detection (IASD), and Incompatible Visual Question Detection (IVQD). To deeply investigate the UPD problem, extensive experim…

"Coffee grounds might be the answer to agricultural contamination: Here’s how"
#Coffee #Agriculture

Could old coffee grounds help clean up contaminated water?
An estimated six million tonnes of coffee grounds are discarded annually. What if they could be put to good use?

Who do you think will pay $30-50B for US TikTok?
Google / Facebook / Amazon
Microsoft / Apple
A Republican-approved coalition
Those are all the same answer

Making Them Ask and Answer: Jailbreaking Large Language Models in Few Queries via Disguise and Reconstruction
Tong Liu, Yingjie Zhang, Zhe Zhao, Yinpeng Dong, Guozhu Meng, Kai Chen
https://arxiv.org/abs/2402.18104

Making Them Ask and Answer: Jailbreaking Large Language Models in Few Queries via Disguise and Reconstruction
In recent years, large language models (LLMs) have demonstrated notable success across various tasks, but the trustworthiness of LLMs is still an open problem. One specific threat is the potential to generate toxic or harmful responses. Attackers can craft adversarial prompts that induce harmful responses from LLMs. In this work, we pioneer a theoretical foundation in LLMs security by identifying bias vulnerabilities within the safety fine-tuning and design a black-box jailbreak method named DR…

Cw: discussion of #transphobia; JKR; politics and power
Someone in a Facebook group I'm in asked:
"What the f☆☆k has JK Rowling got to do with removing the appalling Zionist Keir Starmer & the rest of our neoliberal NATO supporting warmongering parliamentary representatives from power?
Asking for a very angry feminist socialist lesbian disabled 70 year old."
I thought this was actually a deceptively good question, so I thought I'd share my answer here...
The thing to understand is that trans rights and trans healthcare are being used as a 'wedge issue' by the right wing: something to divide and distract us while they get on with their usual business of plundering everybody.
Rowling, as a billionaire with a huge platform and a great deal of time on her hands, has done more than practically anybody else to drive this wedge deeper.
Like every scapegoating campaign used by the right - against gay people, immigrants, minorities, benefits claimants and so on - the anti-trans moral panic has serious, material consequences for its targets.
The ones who benefit from these hate campaigns (and often fund them, directly or indirectly) are largely indifferent to the harm caused; the distraction and division are the main points, but stoking fear and hatred are time-worn methods of achieving those.
Hope this helps!

#Framed #720
🎥 🟥 🟥 🟩 ⬛ ⬛ ⬛
https://framed.wtf

His final message on Facebook read, “Many of us like to ask ourselves, ‘What would I do if I was alive during slavery? Or the Jim Crow South? Or apartheid? What would I do if my country was committing genocide?’ The answer is, you’re doing it. Right now.”
https://www.newy…

Aaron Bushnell’s Act of Political Despair
Masha Gessen asks what it means for an American to self-immolate, as Aaron Bushnell did to protest U.S. support for Israel’s war in Gaza.

Even in the #StarTrek universe, alien encounters are not always peaceful. And for today's #TrekTriviaTuesday we'll look at a particular devastating one.
As always no googling and no spoiling the answer for others. Please boost after voting! :BoostOK:
Vote will run for 24h, …

Klasse Thema: #StefanGeißler von #Kairntech bei der @… in seinem #Vortrag

#potd "My" turn-right-on-red (if you're on a #bicycle) sign was installed! I suggested the location last year, and got a very quick answer from MA46 that they will install it in spring.

The Reflection Coefficient of a Reflectionless Kink
Jarah Evslin, Hui Liu
https://arxiv.org/abs/2402.17968 https://arxiv.org/pdf/2402…

The Reflection Coefficient of a Reflectionless Kink
Classically, reflectionless kinks transmit all incident radiation. Recently, we have used an analyticity argument together with a solution of the Lippmann-Schwinger equation to write down the leading quantum correction to the reflection probability. The argument was fast but rather indirect. In the present paper, we calculate the reflection coefficient and probability by methodically grinding through the Schrodinger picture time evolution. We find the same answer. This answer contains contribut…

The Boston-area Covid sewage RNA counts have not improved in a year
I've been copying the Covid RNA-in-sewage measurements from the MWRA's website for a few years now, mostly so I could replot it on a log scale, but also to answer (vaguely, sort of, with caveats) questions about how we are doing.  The short answer is we are not doing any better than we were last year, and have been behind more or less since mid-December.

The Boston-area Covid sewage RNA counts have not improved in a year
I’ve been copying the Covid RNA-in-sewage measurements from the MWRA’s website for a few years now, mostly so I could replot it on a log scale, but also to answer (vaguely, sort of, wit…

So I've tried out #Bing's #AI copilot to answer me some #PySide questions. It's great that I can talk to it in German since I usually need to search in English on StackOverflow. I've asked it which …

Mike Tomlin excited about Steelers’ mindset: ‘We as a collective have some questions to answer’ https://www.nfl.com/news/mike-tomlin-excited-about-steelers-mindset-we-as-a-collective-have-some-questions-to-answer

Mike Tomlin excited about Steelers’ mindset: ‘We as a collective have some questions to answer’
Steelers head coach Mike Tomlin is encouraged by the prove-it mindset of offensive coordinator Arthur Smith and quarterbacks Russell Wilson and Justin Fields. Pittsburgh added a number of value picks in the draft to help them achieve such goals.

STOP FUELLING WAR: It’s Time for Action!: https://benborges.xyz/2024/04/01/stop-fuelling-war.html

#Wordle 985 4/6*
⬜🟨⬜⬜⬜ <1% of 255,044 (272)
⬜🟨🟩⬜⬜ 3% of 34 (15)
⬜⬜🟩🟨🟩 24% of 86 (2)
🟩🟩🟩🟩🟩
WordleBot
Skill 88/99
Luck 52/99
A good day, No real struggles to find words to test new letters and new placements leading right to the answer

How can we evaluate the factual accuracy of long answers from LLMs? Researchers from DeepMind / Stanford demonstrate a strategy that uses LLMs search to assess factuality: it's more accurate than human evaluation and 20x cheaper. h/t Marc Lanctot on Threads arxiv.org/abs/2403.18802

Note to self: If you are a journalist and want to ask a company one critical, important question, and maybe a few others, it's probably a better idea to skip the few others, or ask them in a later mail. Avoids that they can play games and only answer the lesser important questions and just skip the relevant one.

Multi-hop Question Answering over Knowledge Graphs using Large Language Models
Abir Chakraborty
https://arxiv.org/abs/2404.19234 https://

Multi-hop Question Answering over Knowledge Graphs using Large Language Models
Knowledge graphs (KGs) are large datasets with specific structures representing large knowledge bases (KB) where each node represents a key entity and relations amongst them are typed edges. Natural language queries formed to extract information from a KB entail starting from specific nodes and reasoning over multiple edges of the corresponding KG to arrive at the correct set of answer nodes. Traditional approaches of question answering on KG are based on (a) semantic parsing (SP), where a logi…

Despite the Ukraine aid vote, the neo-isolationist threat still looms
Is America back for good or only for now?
There is no way to answer that question with any degree of confidence.
And that, in turn, should give U.S. allies pause about whether they can still count on the United States.

While overwhelming majorities of both houses wound up backing aid to Ukraine,
narrow majorities of Republicans opposed the bill in the House
and in the first Senate vo…

Why U.S. aid to Ukraine is still in jeopardy
U.S. allies may be relieved by Congress’s action, but they’d be wise to keep strengthening their defenses, given Donald Trump’s possible election next fall.

Cowboys' Stephen Jones hinted at Leighton Vander Esch's outlook https://cowboyswire.usatoday.com/2024/02/28/stephen-jones-leighton-vander-esch-update/

Cowboys' Stephen Jones hinted at Leighton Vander Esch's outlook
From @ToddBrock24f7: Jones didn’t directly address the linebacker’s injury status, but his non-answer did not paint an optimistic picture of an imminent return.

When to Retrieve: Teaching LLMs to Utilize Information Retrieval Effectively
Tiziano Labruna, Jon Ander Campos, Gorka Azkune
https://arxiv.org/abs/2404.19705 https://arxiv.org/pdf/2404.19705
arXiv:2404.19705v1 Announce Type: new
Abstract: In this paper, we demonstrate how Large Language Models (LLMs) can effectively learn to use an off-the-shelf information retrieval (IR) system specifically when additional context is required to answer a given question. Given the performance of IR systems, the optimal strategy for question answering does not always entail external information retrieval; rather, it often involves leveraging the parametric memory of the LLM itself. Prior research has identified this phenomenon in the PopQA dataset, wherein the most popular questions are effectively addressed using the LLM's parametric memory, while less popular ones require IR system usage. Following this, we propose a tailored training approach for LLMs, leveraging existing open-domain question answering datasets. Here, LLMs are trained to generate a special token, , when they do not know the answer to a question. Our evaluation of the Adaptive Retrieval LLM (Adapt-LLM) on the PopQA dataset showcases improvements over the same LLM under three configurations: (i) retrieving information for all the questions, (ii) using always the parametric memory of the LLM, and (iii) using a popularity threshold to decide when to use a retriever. Through our analysis, we demonstrate that Adapt-LLM is able to generate the token when it determines that it does not know how to answer a question, indicating the need for IR, while it achieves notably high accuracy levels when it chooses to rely only on its parametric memory.

surprisingly good answers from Adlah Chisti, E.J. Jones and Chyanne Chen, all calling for protected bike lanes!
be nice if the current D11 supervisor who's running for mayor would give as good of an answer as those three. #sfPol

Meet the District 11 candidates: ‘Do you think Vision Zero has failed SF?’
阅读中文版 Welcome back to “Meet the Candidates” series for District 11, where we ask each candidate to answer one question per week leading up to the

Correct! There IS no correct answer. Thanks for the help, though, everyone. I'll bear the best answers in mind... https://mas.to/@outer/111983284803907909

Yeah, you're really gonna see which companies are just gonna allow the AI to scrape all their stuff now. I'm a copyleft/creative commons kinda guy. But if you have art that you don't want stolen, the answer is simple.
MAKE YOUR OWN WEBSITE and put your art there (edit: and use that Glaze type of stuff on your art that wrecks AI, just to be sure)! Neocities is SO easy to set up! Or your own domain and hosting via porkbun, GoDaddy (non-WordPress) - anything at all other than …

This https://arxiv.org/abs/2308.00299 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_mat…

Full Souslin trees at small cardinals
A $κ$-tree is said to be full if each of its limit levels omits no more than one potential branch. Kunen asked whether a full $κ$-Souslin tree may consistently exist. Shelah gave an affirmative answer of height a strong limit Mahlo cardinal. Here, it is shown that these trees may consistently exist at small cardinals. Indeed, there can be $\aleph_3$ many full $\aleph_2$-trees such that the product of any countably many of them is an $\aleph_2$-Souslin tree.

*I bought a donut. And they gave me a receipt for the donut.*
Do you remember when you'd participate in an online community and your reward for doing so was.... participation in that community?
*I don’t need a receipt for a donut. I’ll just give you the money. You give me the donut. End of transaction.*
Now you post in some community or forum and you get participation awards.
*We don’t need to bring ink and paper into this.*
I just got an award.
1/…

Meta Refuses to Answer Questions on Gaza Censorship, Say Sens. Warren and Sanders #Meta

Defensive line draft prospects give praise to Maxx Crosby as one of top players in the league https://www.raiders.com/news/maxx-crosby-defensive-line-combine-2024-draft-prospects-give-praise

Defensive line draft prospects give praise to Maxx Crosby as one of top players in the league
When a few 2024 Combine participants were asked who their idol is in the league, there was one common Silver and Black answer.

Lattices in rigid analytic representations
Andrea Conti, Emiliano Torti
https://arxiv.org/abs/2403.20232 https://arxiv.org/pdf/2403.2…

Lattices in rigid analytic representations
For a profinite group $G$ and a rigid analytic space $X$, we study when an $\mathcal O_X(X)$-linear representation $V$ of $G$ admits a lattice, i.e. an $\mathcal O_{\mathcal X}(\mathcal X)$-linear model for a formal model $\mathcal X$ of $X$ in the sense of Berthelot. We give a positive answer, under mild assumptions, when $X$ is a "wide open" space. Via such a result, we are able to describe explicit subdomains of $X$ over which $V$ is constant after reduction modulo a power of $p$. As an appl…

January 4, 2024 NYT crossword puzzle spoiler and petty gripe
NYT crosswords are usually so precise, but this one has a mistake, which I think is the first one I ever found after doing thousands of them over the years.
The clue is "36:1, for snake-eyes" and the correct answer is "ODDS". But the odds for snake-eyes are 1:35. Argh!
#NYTCrossword #Math #Probability

This https://arxiv.org/abs/2301.02206 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csSE_…

Lessons from Formally Verified Deployed Software Systems (Extended version)
The technology of formal software verification has made spectacular advances, but how much does it actually benefit the development of practical software? Considerable disagreement remains about the practicality of building systems with mechanically-checked proofs of correctness. Is this prospect confined to a few expensive, life-critical projects, or can the idea be applied to a wide segment of the software industry? To help answer this question, the present survey examines a range of projects…

This https://arxiv.org/abs/2306.09852 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csRO_…

Actor-Critic Model Predictive Control
An open research question in robotics is how to combine the benefits of model-free reinforcement learning (RL) - known for its strong task performance and flexibility in optimizing general reward formulations - with the robustness and online replanning capabilities of model predictive control (MPC). This paper provides an answer by introducing a new framework called Actor-Critic Model Predictive Control. The key idea is to embed a differentiable MPC within an actor-critic RL framework. The prop…

@… I think “it’s complicated” is a fine answer! (though not reflected in the poll choices 😅)

Dear LazyWeb, Bit of a long shot here. I'm replacing a bog-standard 5T disk in my bog-standard Synology DS416j and seem to have lost the package of screws that came with the thing. Having trouble figuring out exactly what the right kind of replacement screws is required to fasten the disk in the tray. Thought I had, but it's not really locked down, and vibrates. Anyone know the right answer?

This https://arxiv.org/abs/2309.17155 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_qu…

Immersed figure-8 annuli and anyons
Immersion (i.e., local embedding) is relevant to the physics of topologically ordered phases through entanglement bootstrap. An annulus can immerse in a disk or a sphere as a ``figure-8", which cannot be smoothly deformed to an embedded annulus. We investigate a simple problem: is there an Abelian state on the immersed figure-8 annulus, locally indistinguishable from the ground state of the background physical system? We show that if the answer is affirmative, a strong sense of isomorphism must…

inploid: Inploid: an online social Q&A platform
Inploid is a social question & answer website in Turkish. Users can follow others and see their questions and answers on the main page. Each user is associated with a reputability score which is influenced by feedback of others about questions and answers of the user. Each user can also specify interest in topics. The data is crawled in June 2017 and consist of 39,749 nodes and 57,276 directed links between them. In addition, for …

This https://arxiv.org/abs/2402.11315 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_grqc_…

Rigidity aspects of Penrose's singularity theorem
In this paper, we study rigidity aspects of Penrose's singularity theorem. Specifically, we aim to answer the following question: if a spacetime satisfies the hypotheses of Penrose's singularity theorem except with weakly trapped surfaces instead of trapped surfaces, then what can be said about the global spacetime structure if the spacetime is null geodesically complete? In this setting, we show that we obtain a foliation of MOTS which generate totally geodesic null hypersurfaces. Depending on…

Out-of-Distribution Detection using Maximum Entropy Coding
Mojtaba Abolfazli, Mohammad Zaeri Amirani, Anders H{\o}st-Madsen, June Zhang, Andras Bratincsak
https://arxiv.org/abs/2404.17023

Out-of-Distribution Detection using Maximum Entropy Coding
Given a default distribution $P$ and a set of test data $x^M=\{x_1,x_2,\ldots,x_M\}$ this paper seeks to answer the question if it was likely that $x^M$ was generated by $P$. For discrete distributions, the definitive answer is in principle given by Kolmogorov-Martin-Löf randomness. In this paper we seek to generalize this to continuous distributions. We consider a set of statistics $T_1(x^M),T_2(x^M),\ldots$. To each statistic we associate its maximum entropy distribution and with this a univ…

They did it again...
The next weird device to run Doom? Your lawnmower
https://www.rockpapershotgun.com/the-next-weird-device-to-run-doom-your-lawnmower?utm_source=feed&utm_medium=…

The next weird device to run Doom? Your lawnmower
"Can it run Doom?" is my favourite question, and lawnmower company Husqvarna have a wonderfully weird answer.

This is a fascinating read. Unsurprisingly, the answer is “capitalism” with the one-size-fits-all mentality. Lowest common denominator all the way down. https://hachyderm.io/@skinnylatte/112001917313689580

This https://arxiv.org/abs/2307.15852 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_mat…

Dimensionless Policies based on the Buckingham $π$ Theorem: Is This a Good Way to Generalize Numerical Results?
The answer to the question posed in the title is yes if the context (the list of variables defining the motion control problem) is dimensionally similar. This article explores the use of the Buckingham $π$ theorem as a tool to encode the control policies of physical systems into a more generic form of knowledge that can be reused in various situations. This approach can be interpreted as enforcing invariance to the scaling of the fundamental units in an algorithm learning a control policy. Fir…

This https://arxiv.org/abs/2308.16885 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_mat…

Chromatic number of spacetime
We observe that an old theorem of Graham implies that for any positive integer $r$, there exists some positive integer $T(r)$ such that every $r$-colouring of $\mathbb{Z}^2$ contains a monochromatic pair of points $(x,y),(x',y')$ with $(x-x')^2 - (y-y')^2 = T(r)$. By scaling, this implies that every finite colouring of $\mathbb{R}^2$ contains a monochromatic pair of points $(x,y),(x',y')$ with $(x-x')^2 - (y-y')^2 = 1$, which answer a problem of Kosheleva and Kreinovich on a pseudo-Euclidean an…

Quantitative Results on Symplectic Barriers
Pazit Haim-Kislev, Richard Hind, Yaron Ostrover
https://arxiv.org/abs/2404.19396 https://arxiv.org/pdf/2404.19396
arXiv:2404.19396v1 Announce Type: new
Abstract: In this paper we present some quantitative results concerning symplectic barriers. In particular, we answer a question raised by Sackel, Song, Varolgunes, and Zhu regarding the symplectic size of the $2n$-dimensional Euclidean ball with a codimension-two linear subspace removed.

every day I wake up and think "what can I buy that will bring me happiness today" and the answer is always the same: nothing. Donating, on the other hand, is incredibly satisfying.

“How to explain their willingness to support Trump despite the risk he poses to our system of government? The answer is not rapidly changing technology, widening inequality, unsuccessful foreign policies or unrest on university campuses but something much deeper and more fundamental. It is what the Founders worried about and Abraham Lincoln warned about: a decline in what they called public virtue.”
#GiftArticle

We have a radical democracy. Will Trump voters destroy it?
The Founding Fathers feared that Americans would someday reject their many freedoms. That’s what the 2024 election is really about.

New frontier in computer science paper titles just dropped. Sure, sure, people have likely used variants of "to be or not to be", but how many people quote from Tamburlaine? Finally my humanities education comes in handy! [From https://www.nytimes.com/2024/03/28/books/r

Like My Book Title? Thanks, I Borrowed It.
Literary allusions are everywhere. What are they good for?

You know what's youthful? Answer: the socialists of #NHPolitics

This https://arxiv.org/abs/2307.14690 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_mat…

A Kodaira type conjecture on almost complex 4 manifolds
Not long ago, Cirici and Wilson defined a Dolbeault cohomology on almost complex manifolds to answer Hirzebruch's problem. In this paper, we define a refined Dolbeault cohomology on almost complex manifolds. We show that the condition $\tilde h^{1,0}=\tilde h^{0,1}$ implies a symplectic structure on a compact almost complex $4$ manifold, where $\tilde h^{1,0}$ and $\tilde h^{0,1}$ are the dimensions of the refined Dolbeault cohomology groups with bi-degrees $(1,0)$ and $(0,1)$ respectively. Com…

Corpus-Steered Query Expansion with Large Language Models
Yibin Lei, Yu Cao, Tianyi Zhou, Tao Shen, Andrew Yates
https://arxiv.org/abs/2402.18031 https://<…

Corpus-Steered Query Expansion with Large Language Models
Recent studies demonstrate that query expansions generated by large language models (LLMs) can considerably enhance information retrieval systems by generating hypothetical documents that answer the queries as expansions. However, challenges arise from misalignments between the expansions and the retrieval corpus, resulting in issues like hallucinations and outdated information due to the limited intrinsic knowledge of LLMs. Inspired by Pseudo Relevance Feedback (PRF), we introduce Corpus-Steer…

surprisingly good answers from Adlah Chisti, E.J. Jones and Chyanne Chen, all calling for protected bike lanes!
be nice if the current D11 supervisor who's running for mayor would give as good of an answer as those three. #sfPol

Meet the District 11 candidates: ‘Do you think Vision Zero has failed SF?’
阅读中文版 Welcome back to “Meet the Candidates” series for District 11, where we ask each candidate to answer one question per week leading up to the

Stop nagging! When I tell you no, then the answer is NO.

I don’t think Israelis or the Biden administration fully appreciate the rage that is bubbling up around the world, fueled by social media and TV footage, over the deaths of so many thousands of Palestinian civilians, particularly children, with U.S.-supplied weapons in Israel’s war in Gaza.
Hamas has much to answer for in triggering this human tragedy, but Israel and the U.S. are seen as driving events now and getting most of the blame.
That such anger is boiling over in the Ara…

Opinion | Israel Is Losing Its Greatest Asset: Acceptance
Israel’s very legitimacy is being undermined by the enormous casualties in Gaza and its government’s failure to provide a credible postwar plan.

"What determines the political preferences of your #AIchatbot?
Conservatives complain that many commercially available A.I. bots exhibit a persistent liberal bias. Elon Musk built Grok as an alternative language model after grumbling about ChatGPT being a “woke” A.I.
Liberals notice that A.I. output is often—in every sense—insufficiently diverse, because models learn from correlatio…

Opinion | Quiz: How Do Your Politics Stack Up Against ChatGPT’s?
We may soon rely a lot on A.I. chatbots, so keeping an eye on their political leanings is becoming more and more important.

I REALLY wish people would stop saying
“please stop using x”
If their answer to someone asking “what should I use instead?”
Is: “I dunno, something else.”
Just fucking stop.
(╯°□°）╯︵ ┻━┻
What do I suggest you do instead?
“Shut the fuck up”.
I will also accept “fuck I hate x, I wish there was an alternative to it.”

One of the great curses in technology is that, if you’ve ever built a content management system, especially a blogging tool, you can never stop building better ones in your head, even though the actual products in reality never get any better. Here @… falls down the rabbit hole again.

42 ... is not the answer to everything but the hours that have passed since the last (and only moderately detailled) update on the status of and crucial communications effort with #Odysseus on the Moon - not good given the total lifetime of the lander of one week in the best case: almost three of these days have now passed. To ease our nerves here are some #snowdrop (Galanthus nivalis) pictures I just took in my neighborhood in #Bochum, Germany ...

Thermodynamic & algorithmic efficiency in heterogeneous networks
Davide Cipollini, Lambert Schomaker
https://arxiv.org/abs/2404.17829 https://

Thermodynamic & algorithmic efficiency in heterogeneous networks
Why complex structures emerge in a real world environment is a fascinating question which has not found any fully satisfactory answer at the point of today. At the border between statistical physics, machine learning and network theory, we provide some computational empirical evidence indicating that the emergence of complex structures in a real world environment is the macroscopic thermodynamic result of the optimal trade-off between the ability to encode information about the environment in t…

#Wordle 1,046 5/6*
⬜🟩⬜⬜⬜ <1% of 228,744 (113)
⬜🟩🟩🟨⬜ 3% of 110 (5)
🟩🟩🟩⬜⬜ 13% of 24 (3)
🟩🟩🟩⬜⬜ 43% of 1,972 (1)
🟩🟩🟩🟩🟩
WordleBot
Skill 88/99
Luck 43/99
On my 3rd guess I had a lot of ideas. I went w/ one on the sort of riskier side, but totally possible. Then something else I thought was maybe better on guess 4, and then I got worried, and lucky!
Seems I had thought of 4 of the 5 choices for guess 3! The answer however was the 5th choice which I got as the sole remaining option on my 5th guess

Every time I see folk complaining about the quality of #AI output I remind myself that there is a certain base level of intellect needed to meaningfully discuss any answer in an informed way, and so the opinion is coming from someone who is usually of above average intelligence and can evaluate answers effectively.
I balance that by thinking about the people in bars and on social media who will te…

Gestern ging ein Meme durch die Runde, wie sich das Hobby Vögelbeobachtung an einen heran schleicht.
Heute habe ich #Merlin von Cornell ausprobiert und bin sehr begeistert: offline, live Ansicht aller erkannten Vögel und ohne Account nutzbar.
Hatte #BirdNet getestet, aber Merlin ist deutlich coole…

Merlin Bird ID - Home
Bird ID Wizard—Step-by-step Answer three simple questions about a bird you are trying to identify and Merlin will give you a list of possible matches. Merlin offers quick identification help for all levels of bird watchers and outdoor enthusiasts to help you learn about the birds in any country i

The NTSB youtube channel is very nice; you get the raw press briefing videos, but also a lot of 'B-Roll' uncommented, for example today video of them just examining the ship, and plugging a USB stick into it (well, it's VDR).
But heck, the press conferences are pretty good - Jennifer Homendy, the NTSB chair, has run ones on Plane crashes, train derailments, and now the ship/bridge collision and can give a pretty good answer before handing off to specialists.

An interesting interpretation I’ve read somewhere of forty-two being the answer to life, the universe, everything is that, if you add up all the sides of a die, you get twenty-one. So, if you have two, that would be forty-two.
Hence, the answer to life, the universe, everything is 2 die.

If your answer is micro generation, no man is an island
And grids that are islands struggle
https://www.youtube.com/watch?v=bbECmVdyWlQ
Teamwork makes the dream work

“I am in shock that a lawyer stood in the U.S. Supreme Court and said that a president could assassinate his political opponent and it would be immune as ‘an official act.’ I am in despair that several Justices seemed to think this answer made perfect sense.”
https://newrepublic.…

The Court Just Sealed Everyone’s Fate, Including Its Own
The justices seem to think that the power they apparently just handed Donald Trump can’t be used against them someday. Right.

Correct! There IS no correct answer. Thanks for the help, though, everyone. I'll bear the best answers in mind... https://mas.to/@outer/111983284803907909

Raiders have met with best cornerback at NFL Combine https://www.yardbarker.com/nfl/articles/raiders_have_met_with_best_cornerback_at_nfl_combine/s1_17313_40036680

Raiders have met with best cornerback at NFL Combine
During the 2024 NFL Draft Combine, the Las Vegas Raiders and every other team are in Indianapolis to watch the players perform drills, get their measurables done, and answer questions in front of media and that the teams ask. One of the Raiders' biggest needs, outside of quarterback, of course, is cornerback, in which the Raiders will have plenty of guys to choose from. In this case, they could end up with the best cornerback in the draft at pick No. 13 in, Terrion Arnold, who is amazing everyo…

January 4, 2024 NYT crossword puzzle spoiler and petty gripe
NYT crosswords are usually so precise, but this one has a mistake, which I think is the first one I ever found after doing thousands of them over the years.
The clue is "36:1, for snake-eyes" and the correct answer is "ODDS". But the odds for snake-eyes are 1:35. Argh!
#NYTCrossword #Math #Probability

Alien languages are a big part of #StarTrek lore. So how about some Klingon history or better etymology for today's #TrekTriviaTuesday.
As always no googling and no spoiling the answer for others. Please boost after voting! :BoostOK:
Vote will run for 24h, then I will post the …

They did it again...
The next weird device to run Doom? Your lawnmower
https://www.rockpapershotgun.com/the-next-weird-device-to-run-doom-your-lawnmower?utm_source=feed&utm_medium=…

The next weird device to run Doom? Your lawnmower
"Can it run Doom?" is my favourite question, and lawnmower company Husqvarna have a wonderfully weird answer.

#Wordle 1,046 5/6*
⬜🟩⬜⬜⬜ <1% of 228,744 (113)
⬜🟩🟩🟨⬜ 3% of 110 (5)
🟩🟩🟩⬜⬜ 13% of 24 (3)
🟩🟩🟩⬜⬜ 43% of 1,972 (1)
🟩🟩🟩🟩🟩
WordleBot
Skill 88/99
Luck 43/99
On my 3rd guess I had a lot of ideas. I went w/ one on the sort of riskier side, but totally possible. Then something else I thought was maybe better on guess 4, and then I got worried, and lucky!
Seems I had thought of 4 of the 5 choices for guess 3! The answer however was the 5th choice which I got as the sole remaining option on my 5th guess

NewsQs: Multi-Source Question Generation for the Inquiring Mind
Alyssa Hwang, Kalpit Dixit, Miguel Ballesteros, Yassine Benajiba, Vittorio Castelli, Markus Dreyer, Mohit Bansal, Kathleen McKeown
https://arxiv.org/abs/2402.18479

NewsQs: Multi-Source Question Generation for the Inquiring Mind
We present NewsQs (news-cues), a dataset that provides question-answer pairs for multiple news documents. To create NewsQs, we augment a traditional multi-document summarization dataset with questions automatically generated by a T5-Large model fine-tuned on FAQ-style news articles from the News On the Web corpus. We show that fine-tuning a model with control codes produces questions that are judged acceptable more often than the same model without them as measured through human evaluation. We …

A Categorization of Complexity Classes for Information Retrieval and Synthesis Using Natural Logic
Gregory Coppola
https://arxiv.org/abs/2402.18566 https:/…

A Categorization of Complexity Classes for Information Retrieval and Synthesis Using Natural Logic
Given the emergent reasoning abilities of large language models, information retrieval is becoming more complex. Rather than just retrieve a document, modern information retrieval systems advertise that they can synthesize an answer based on potentially many different documents, conflicting data sources, and using reasoning. But, different kinds of questions have different answers, and different answers have different complexities. In this paper, we introduce a novel framework for analyzing the…

Mailbag: Is free agent RB the answer? https://www.dallascowboys.com/news/mailbag-is-free-agent-rb-the-answer

Mailbag: Is free agent RB the answer?
Fans are seeing headlines about bringing in a free agent like Derrick Henry, but I have been trying to recall the last free agent running back who signed elsewhere and continued to have the same amount of success. Is history enough to avoid free agency to fill the hole? 

"What determines the political preferences of your #AIchatbot?
Conservatives complain that many commercially available A.I. bots exhibit a persistent liberal bias. Elon Musk built Grok as an alternative language model after grumbling about ChatGPT being a “woke” A.I.
Liberals notice that A.I. output is often—in every sense—insufficiently diverse, because models learn from correlatio…

Opinion | Quiz: How Do Your Politics Stack Up Against ChatGPT’s?
We may soon rely a lot on A.I. chatbots, so keeping an eye on their political leanings is becoming more and more important.

Falcons coach Raheem Morris explains how the team will find its best fit at quarterback

https://www.cbssports.com/nfl/news/falcons-…

Falcons coach Raheem Morris explains how the team will find its best fit at quarterback
The Falcons are looking for a new answer under center after 2023's struggles

The NTSB youtube channel is very nice; you get the raw press briefing videos, but also a lot of 'B-Roll' uncommented, for example today video of them just examining the ship, and plugging a USB stick into it (well, it's VDR).
But heck, the press conferences are pretty good - Jennifer Homendy, the NTSB chair, has run ones on Plane crashes, train derailments, and now the ship/bridge collision and can give a pretty good answer before handing off to specialists.

Finally we have a CLI to definitively answer whether something is a web site or a web app: https://github.com/Y2Z/monolith
See if you can tell which is which (bonus points if you can guess which HTML file is 9.6MB and which is 703KB)

qa_user: User interactions on Q&A websites (2016)
Networks of interactions among users from four online Q&A sites: Stack Overflow, Math Overflow, Super User, and Ask Ubuntu. A directed edge (i,j) indicates a user i responded to user j's post. Edges are timestamped. For each Q&A site, four differently defined networks are provided, based on the definition of an edge: (i) a user answered a question, (ii) a user commented on a question, (iii) a user commented on an answer…

If your answer is micro generation, no man is an island
And grids that are islands struggle
https://www.youtube.com/watch?v=bbECmVdyWlQ
Teamwork makes the dream work

The First Place Solution of WSDM Cup 2024: Leveraging Large Language Models for Conversational Multi-Doc QA
Yiming Li, Zhao Zhang
https://arxiv.org/abs/2402.18385

The First Place Solution of WSDM Cup 2024: Leveraging Large Language Models for Conversational Multi-Doc QA
Conversational multi-doc question answering aims to answer specific questions based on the retrieved documents as well as the contextual conversations. In this paper, we introduce our winning approach for the "Conversational Multi-Doc QA" challenge in WSDM Cup 2024, which exploits the superior natural language understanding and generation capability of Large Language Models (LLMs). We first adapt LLMs to the task, then devise a hybrid training strategy to make the most of in-domain unlabeled da…

“I am in shock that a lawyer stood in the U.S. Supreme Court and said that a president could assassinate his political opponent and it would be immune as ‘an official act.’ I am in despair that several Justices seemed to think this answer made perfect sense.”
https://newrepublic.…

The Court Just Sealed Everyone’s Fate, Including Its Own
The justices seem to think that the power they apparently just handed Donald Trump can’t be used against them someday. Right.

What to expect at MWC 2024: smartphone launches from Xiaomi and others, OnePlus Watch 2, Brain.ai's "app-less phone" and other experimental concepts, and more (Jon Porter/The Verge)
https://www.theverge.com/2024/2/24/24078970/m…

What to expect from MWC 2024
It’ll be hard to evolve today’s smartphone without optimizing thousands or even millions of apps for new designs. Could artificial intelligence offer the answer?

This https://arxiv.org/abs/2312.17269 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCL_…

Conversational Question Answering with Reformulations over Knowledge Graph
Conversational question answering (convQA) over knowledge graphs (KGs) involves answering multi-turn natural language questions about information contained in a KG. State-of-the-art methods of ConvQA often struggle with inexplicit question-answer pairs. These inputs are easy for human beings to understand given a conversation history, but hard for a machine to interpret, which can degrade ConvQA performance. To address this problem, we propose a reinforcement learning (RL) based model, CornNet,…

Generate then Retrieve: Conversational Response Retrieval Using LLMs as Answer and Query Generators
Zahra Abbasiantaeb, Mohammad Aliannejadi
https://arxiv.org/abs/2403.19302

Generate then Retrieve: Conversational Response Retrieval Using LLMs as Answer and Query Generators
CIS is a prominent area in IR that focuses on developing interactive knowledge assistants. These systems must adeptly comprehend the user's information requirements within the conversational context and retrieve the relevant information. To this aim, the existing approaches model the user's information needs with one query called rewritten query and use this query for passage retrieval. In this paper, we propose three different methods for generating multiple queries to enhance the retrieval. I…

Iterative Reasoning Preference Optimization
Richard Yuanzhe Pang, Weizhe Yuan, Kyunghyun Cho, He He, Sainbayar Sukhbaatar, Jason Weston
https://arxiv.org/abs/2404.19733 https://arxiv.org/pdf/2404.19733
arXiv:2404.19733v1 Announce Type: new
Abstract: Iterative preference optimization methods have recently been shown to perform well for general instruction tuning tasks, but typically make little improvement on reasoning tasks (Yuan et al., 2024, Chen et al., 2024). In this work we develop an iterative approach that optimizes the preference between competing generated Chain-of-Thought (CoT) candidates by optimizing for winning vs. losing reasoning steps that lead to the correct answer. We train using a modified DPO loss (Rafailov et al., 2023) with an additional negative log-likelihood term, which we find to be crucial. We show reasoning improves across repeated iterations of this scheme. While only relying on examples in the training set, our approach results in increasing accuracy for Llama-2-70B-Chat from 55.6% to 81.6% on GSM8K (and 88.7% with majority voting out of 32 samples), from 12.5% to 20.8% on MATH, and from 77.8% to 86.7% on ARC-Challenge, which outperforms other Llama-2-based models not relying on additionally sourced datasets.

TIGQA:An Expert Annotated Question Answering Dataset in Tigrinya
Hailay Teklehaymanot, Dren Fazlija, Niloy Ganguly, Gourab K. Patro, Wolfgang Nejdl
https://arxiv.org/abs/2404.17194

TIGQA:An Expert Annotated Question Answering Dataset in Tigrinya
The absence of explicitly tailored, accessible annotated datasets for educational purposes presents a notable obstacle for NLP tasks in languages with limited resources.This study initially explores the feasibility of using machine translation (MT) to convert an existing dataset into a Tigrinya dataset in SQuAD format. As a result, we present TIGQA, an expert annotated educational dataset consisting of 2.68K question-answer pairs covering 122 diverse topics such as climate, water, and traffic. …