Negation instructions such as 'do not mention $X$' can paradoxically increase the accessibility of $X$ in human thought, a phenomenon known as ironic rebound. Large language models (LLMs) face the same challenge: suppressing a concept requires internally activating it, which may prime rebound instead of avoidance. We investigated this tension with two experiments. \textbf{(1) Load \& content}: after a negation instruction, we vary distractor text (semantic, syntactic, repetition) and measure rebound strength. \textbf{(2) Polarity separation}: We test whether models distinguish neutral from negative framings of the same concept and whether this separation predicts rebound persistence. Results show that rebound consistently arises immediately after negation and intensifies with longer or semantic distractors, while repetition supports suppression. Stronger polarity separation correlates with more persistent rebound. Together, these findings, complemented by a circuit tracing analysis that identifies sparse middle-layer attention heads amplifying forbidden tokens while early layers suppress, link cognitive predictions of ironic rebound with mechanistic insights into long-context interference. To support future work, we release ReboundBench, a dataset of $5,000$ systematically varied negation prompts designed to probe rebound in LLMs.

The final target point of the presented research leads us to bio - inspired mobile robots, especially those able to reconstruct the natural mobility of gibbons. The principal mode of their locomotion is called brachiation. It consists of swinging from branch to branch for distances of up to 15 m and at speeds up to 50 km/h (Figure 1). We may address the readers to several brachiation techniques and constructions presented in the technical literature [1 - 5]. Seeing several similarities, we may classify the brachi ation robots as a branch of the walking ones (Fig.1a). Each research on the brachiation dynamics is challenging, mainly because of their multitasking: the system's number of degrees of freedom varies during the motion (i.e., we need model a nonlinear time - varying system), unilateral constraints are present (i.e., impact forces can appear) at selected stages of their locomotion, the investigated systems are kinematically or dynamically overactuated.

Two issues of algorithmic collusion are addressed in this paper. First, we show that in a general class of symmetric games, including Prisoner's Dilemma, Bertrand competition, and any (nonlinear) mixture of first and second price auction, only (strict) Nash Equilibrium (NE) is stochastically stable. Therefore, the tacit collusion is driven by failure to learn NE due to insufficient learning, instead of learning some strategies to sustain collusive outcomes. Second, we study how algorithms adapt to collusion in real simulations with insufficient learning. Extensive explorations in early stages and discount factors inflates the Q-value, which interrupts the sequential and alternative price undercut and leads to bilateral rebound. The process is iterated, making the price curves like Edgeworth cycles. When both exploration rate and Q-value decrease, algorithms may bilaterally rebound to relatively high common price level by coincidence, and then get stuck. Finally, we accommodate our reasoning to simulation outcomes in the literature, including optimistic initialization, market design and algorithm design.

Large language models hold significant potential for integrating various data types, such as text documents and database records, for advanced analytics. However, blending text and numerical data presents substantial challenges. LLMs need to process and cross-reference entities and numbers, handle data inconsistencies and redundancies, and develop planning capabilities such as building a working memory for managing complex data queries. In this paper, we introduce four novel tasks centered around sports data analytics to evaluate the numerical reasoning and information fusion capabilities of LLMs. These tasks involve providing LLMs with detailed, play-by-play sports game descriptions, then challenging them with adversarial scenarios such as new game rules, longer durations, scrambled narratives, and analyzing key statistics in game summaries. We conduct extensive experiments on NBA and NFL games to assess the performance of LLMs on these tasks. Our benchmark, SportsMetrics, introduces a new mechanism for assessing LLMs' numerical reasoning and fusion skills.

Jumping and hopping locomotion are efficient means of traversing unstructured rugged terrain with the former being the focus of roboticists. This focus has led to significant performance and understanding in jumping robots but with limited practical applications as they require significant time between jumps to store energy, thus relegating jumping to a secondary role in locomotion. Hopping locomotion, however, can preserve and transfer energy to subsequent hops without long energy storage periods. Therefore, hopping has the potential to be far more energy efficient and agile than jumping. However, to date, only a single untethered hopping robot exists with limited payload and hopping heights (< 1 meter). This is due to the added design and control complexity inherent in the requirements to input energy during dynamic locomotion and control the orientation of the system throughout the hopping cycle, resulting in low energy input and control torques; a redevelopment from basic principles is necessary to advance the capabilities of hopping robots. Here we report hopping robot design principles for efficient and robust systems with high energy input and control torques that are validated through analytical, simulation, and experimental results. The resulting robot (MultiMo-MHR) can hop nearly 4 meters (> 6 times the current state-of-the-art); and is only limited by the impact mechanics and not energy input. The results also directly contradict a recent work that concluded hopping with aerodynamic energy input would be less efficient than flight for hops greater than 0.4 meters.

Large language models (LLMs) have demonstrated an impressive ability to synthesize plausible and fluent text. However they remain vulnerable to hallucinations, and thus their outputs generally require manual human verification for high-stakes applications, which can be timeconsuming and difficult. This paper proposes symbolically grounded generation (SymGen) as a simple approach for enabling easier validation of an LLM's output. SymGen prompts an LLM to interleave its regular output text with explicit symbolic references to fields present in some conditioning data (e.g., a table in JSON format). The references can be used to display the provenance of different spans of text in the generation, reducing the effort required for manual verification. Across data-to-text and question answering experiments, we find that Figure 1: Compare a standard LLM-generated (A) with LLMs are able to directly output text that makes a SymGen (B, ours) description of a basketball game, use of symbolic references while maintaining based on statistics about it.

We study a planning problem based on Plotting, a tile-matching puzzle video game published by Taito in 1989. The objective of this game is to remove a target number of coloured blocks from a grid by sequentially shooting blocks into the grid. Plotting features complex transitions after every shot: various blocks are affected directly, while others can be indirectly affected by gravity. We highlight the challenges of modelling Plotting with PDDL and of solving it with a grounding-based state-of-the-art planner.

Social media researchers at the Network Contagion Research Institute in Princeton, New Jersey, got a rude awakening early last month. They were roused by 6:30 am phone calls from a colleague warning that Reddit had started blocking the institute's Pushshift service from updating its ongoing archive of every post on the discussion platform. That was a problem for more than just NCRI, because some of Reddit's 50,000 volunteer moderators depend on Pushshift to quickly investigate problem users, and many academics rely on the service. If it went stale, mods, as Reddit calls moderators, would have to work overtime or let more trash content accumulate. Researchers studying online communities would be forced to put projects and doctoral dissertations on ice.

MILAN/LONDON – Italy's UniCredit and France's BNP Paribas were the latest banks to set out their Russian exposures, warning of billions of euros in potential costs from the financial fallout from Moscow's invasion of Ukraine. Banks, insurers and asset managers have been scrambling to distance themselves from Russia and assess their exposures after Moscow was hit with heavy sanctions by the West in the wake of the invasion of Ukraine that began last month. Russia calls its actions in Ukraine a "special operation." BNP Paribas has also cut off its Russia-based workforce from its internal computer systems as it seeks to bolster its defenses against any potential cyberattack, a source with direct knowledge of the matter told Reuters. The French lender is believed to be the first major bank to have excluded staff in Moscow from its IT networks.

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. The Tokyo Olympics men's basketball matchup between the U.S. and France was briefly overshadowed by a robot at halftime. The robot was spotted at the foul line taking almost as long as NBA champion Giannis Antetokounmpo to shoot, but the machine made the basket. The robot was then placed at the top of the three-point arch and shot the ball with as much precision as Golden State Warriors star Stephen Curry.