More questions than answers: Arch Manning likes the disrespect, Big 12's disaster and a CFP race in disarray LaNorris Sellers' acrobatics, late fumble recovery help give South Carolina walk-off win over Tennessee Texas Tech's $30 million roster runs into a buzzsaw as Cincinnati delivers Sorsby revenge in Top-10 upset LSU's Mike the Tiger stalks referee while dressed as Michael Myers, appears to be steered away by staffer Ric Flair shows off eyepatch in update after fall left him hospitalized: 'I will be okay' Exclusive: DOJ reviewing Portland Thorns incident showing women's sports supporters struck, forcibly removed Michael Knowles defends archbishop's White House prayer with Trump Trump weighs pardons after Slotkin's warning to young conservatives Sarkozy's son warns violent protests in France could worsen amid political unrest Enes Kanter Freedom: It's going to be a dunk every time Fraud'czar' calls SNAP fraud the'greatest heist in human history' French President Macron blames'ultra-left' for French riots Kaylee McGhee White: Arguments against voter ID are'completely insulting' Rep Jim Jordan commends'common sense conservative' running in tight Florida district OutKick Human vs. ChatGPT NFL Week 7 betting picks: Human clings to profit as AI claims third straight win Blazin' 5 for Week 7: Chiefs cover, Bengals vs. Ravens, Will the Falcons upset the 49ers? Colin Cowherd predicts the best bets entering Week 7 of the NFL season, including the Cincinnati Bengals, Atlanta Falcons, and Kansas City Chiefs. Apparently, the robot doesn't even have to make money to beat me anymore. ChatGPT finished Week 6 at 2-2-1 and lost 0.16 units. I went 2-3 and lost 1.17 units. That's three straight weeks of GPT getting the better of me, although I'm not sure congratulations on losing less money is quite the victory speech it had in mind. My profit is now 0.09 units. Yes, there's still a plus sign in front of my total, and I intend to keep pointing that out until someone makes me stop.

Major League Baseball's best team can't stop pretending to be the underdogs heading into World Series Golfer distracted by'a nice rack' shanks tee shot in the direction of large bull elk, wisely grabs a new ball BYU QB Bear Bachmeier says Notre Dame'disrespected' him and his family following upset win over Irish Former MLB star Jimmy Rollins was not hacked, he is the co-founder of a women's underwear brand Human vs. ChatGPT NFL Week 6 betting picks: AI goes all-in on totals after another weekly win College football chaos: Historic upsets, Arch Manning's fiasco, clock controversy and CFP ramifications arise Hostile Valkyries fans verbally attacked women's sports activists at WNBA Finals Game 1 in San Francisco Data center CEO reveals what's behind the industry's intense security Letitia James is the most'political' AG in the country: Jonathan Turley Rep Jeff Van Drew calls out Obama's defense of Ilhan Omar Netanyahu says Israel'never had an ally' like Trump, details IDF's military strategy Netanyahu opens up about'special relationship' with Trump Arizona's 18-16 win leaves Los Angeles facing questions that extend well beyond a disputed personal foul Former NFL QB Sean Salisbury tells OutKick that officials should answer questions immediately after games, be held accountable for obvious mistakes and stop escalating confrontations with players. The Los Angeles Rams have another reason to complain about NFL officiating. They also have a 2-4 record after losing at home to an Arizona Cardinals team that entered Sunday with one win. I'd be a lot more concerned about the second part. This page may contain affiliate links to legal sports betting partners.

Fever general manager Amber Cox calls out'attacks' on the fan base in season-ending press conference Hall of Famer praises Dan Lanning's'perfect' response after Dante Moore was carted off following vicious hit The ladies of Oktoberfest crush draft beers as fall heats up, Ryan Day's message to sick kids & a Strip fight Virginia Tech's controversial missed field goal sparks fiery debate as different angles tell a different story Rams lose Aaron Donald to back soreness for Eagles game, Sean McVay says he's protecting star'from himself' Tennessee fans drank $1.3 million worth of alcohol before, during Texas game attended by president Tomi Lahren says France protests are a'cautionary tale' for the US Rep Mike Lawler: These leaders don't want to hold people accountable for their actions Daylight savings would be an'easy win' for the Senate, President Trump says Blazin' 5 for Week 4: Seahawks blow out Chargers, Can the Cowboys beat the Texans? Colin Cowherd lists his best bets entering Week 4, including the Seattle Seahawks beating the Los Angeles Chargers, and the Dallas Cowboys bouncing back with a win vs. the Houston Texans. The robot finally had a winning week. Unfortunately for the robot, I had a better one . I went 4-1 in Week 3 and added 3.06 units, bringing my season record to 9-6 and putting me in the black at +2.82 units. GPT managed its first winning card, going 3-2 for a profit of 0.89 units.

Large language models (LLMs) are increasingly used for data generation. However, creating evaluation benchmarks raises the bar for this emerging paradigm. Benchmarks must target specific phenomena, penalize exploiting shortcuts, and be challenging. Through two case studies, we investigate whether LLMs can meet these demands by generating reasoning over-text benchmarks and comparing them to those created through careful crowdsourcing. Specifically, we evaluate both the validity and difficulty of LLM-generated versions of two high-quality reading comprehension datasets: CondaQA, which evaluates reasoning about negation, and DROP, which targets reasoning about quantities. We find that prompting LLMs can produce variants of these datasets that are often valid according to the annotation guidelines, at a fraction of the cost of the original crowdsourcing effort. However, we show that they are less challenging for LLMs than their human-authored counterparts. This finding sheds light on what may have been lost by generating evaluation data with LLMs, and calls for critically reassessing the immediate use of this increasingly prevalent approach to benchmark creation.

Given a sequence of tokens generated by a language model, we may want to identify the preceding tokens that influence the model to generate this sequence. Performing such token attribution is expensive; a common approach is to ablate preceding tokens and directly measure their effects. To reduce the cost of token attribution, we revisit attention weights as a heuristic for how a language model uses previous tokens. Naive approaches to attribute model behavior with attention (e.g., averaging attention weights across attention heads to estimate a token's influence) have been found to be unreliable. To attain faithful attributions, we propose treating the attention weights of different attention heads as features. This way, we can learn how to effectively leverage attention weights for attribution (using signal from ablations). Our resulting method, Attribution with Attention (AT2), reliably performs on par with approaches that involve many ablations, while being significantly more efficient. To showcase the utility of AT2, we use it to prune less important parts of a provided context in a question answering setting, improving answer quality. We provide code for AT2 at https://github.com/MadryLab/AT2 .

Data sanitization in the context of language modeling involves identifying sensitive content, such as personally identifiable information (PII), and redacting them from a dataset corpus. It is a common practice used in natural language processing (NLP) to maintain privacy. Nevertheless, the impact of data sanitization on the language understanding capability of a language model remains less studied. This paper empirically analyzes the effects of data sanitization across several benchmark language-modeling tasks including comprehension question answering (Q&A), entailment, sentiment analysis, and text classification. Our experiments cover a wide spectrum comprising finetuning small-scale language models, to prompting large language models (LLMs), on both original and sanitized datasets, and comparing their performance across the tasks. Interestingly, our results suggest that for some tasks such as sentiment analysis or entailment, the impact of redaction is quite low, typically around 1-5%, while for tasks such as comprehension Q&A there is a big drop of >25% in performance observed in redacted queries as compared to the original. For tasks that have a higher impact, we perform a deeper dive to inspect the presence of task-critical entities. Finally, we investigate correlation between performance and number of redacted entities, and also suggest a strategy to repair an already redacted dataset by means of content-based subsampling. Additional details are available at https://sites.google.com/view/datasan.

Language agents perform complex tasks by using tools to execute each step precisely. However, most existing agents are based on proprietary models or designed to target specific tasks, such as mathematics or multi-hop question answering. We introduce Husky, a holistic, open-source language agent that learns to reason over a unified action space to address a diverse set of complex tasks involving numerical, tabular, and knowledge-based reasoning. Husky iterates between two stages: 1) generating the next action to take towards solving a given task and 2) executing the action using expert models and updating the current solution state. We identify a thorough ontology of actions for addressing complex tasks and curate high-quality data to train expert models for executing these actions. Our experiments show that Husky outperforms prior language agents across 14 evaluation datasets. Moreover, we introduce HuskyQA, a new evaluation set which stress tests language agents for mixed-tool reasoning, with a focus on retrieving missing knowledge and performing numerical reasoning. Despite using 7B models, Husky matches or even exceeds frontier LMs such as GPT-4 on these tasks, showcasing the efficacy of our holistic approach in addressing complex reasoning problems. Our code and models are available at https://github.com/agent-husky/Husky-v1.

In the realm of competitive sports, understanding the performance dynamics of athletes, represented by the age curve (showing progression, peak, and decline), is vital. Our research introduces a novel framework for quantifying age-specific treatment effects, enhancing the granularity of performance trajectory analysis. Firstly, we propose a methodology for estimating the age curve using game-level data, diverging from traditional season-level data approaches, and tackling its inherent complexities with a meta-learner framework that leverages advanced machine learning models. This approach uncovers intricate non-linear patterns missed by existing methods. Secondly, our framework enables the identification of causal effects, allowing for a detailed examination of age curves under various conditions. By defining the Age-Conditioned Treatment Effect (ACTE), we facilitate the exploration of causal relationships regarding treatment impacts at specific ages. Finally, applying this methodology to study the effects of rest days on performance metrics, particularly across different ages, offers valuable insights into load management strategies' effectiveness. Our findings underscore the importance of tailored rest periods, highlighting their positive impact on athlete performance and suggesting a reevaluation of current management practices for optimizing athlete performance.

Fox News Flash top sports headlines are here. Check out what's clicking on Foxnews.com. The first score of the Super Bowl turned out to be historic, but the record didn't stand for long. Kansas City Chiefs kicker Harrison Butker drilled his second field goal of Super Bowl LVIII against the San Francisco 49ers, a 57-yarder that rewrote the record for longest kick in the "Big Game." And that previous record was hit in the first half by 49ers rookie kicker Jake Moody, who hit one home from 56 yards.

Language models are achieving impressive performance on various tasks by aggressively adopting inference-time prompting techniques, such as zero-shot and few-shot prompting. In this work, we introduce EchoPrompt, a simple yet effective approach that prompts the model to rephrase its queries before answering them. EchoPrompt is adapted for both zero-shot and few-shot in-context learning with standard and chain-of-thought prompting. Experimental results show that EchoPrompt yields substantial improvements across all these settings for four families of causal language models. These improvements are observed across various numerical reasoning (e.g. GSM8K, SVAMP), reading comprehension (e.g. DROP), and logical reasoning (e.g. Coin Flipping) tasks. On average, EchoPrompt improves the Zero-shot-CoT performance of code-davinci-002 by 5% in numerical tasks and 13% in reading comprehension tasks. We investigate the factors contributing to EchoPrompt's effectiveness through ablation studies, which reveal that both the original query and the model-generated rephrased version are instrumental in its performance gains. Our empirical results indicate that EchoPrompt is an effective technique that enhances in-context learning performance. We recommend incorporating EchoPrompt into various baseline prompting strategies to achieve performance boosts.