Jailbreaking large language models (LLMs) has emerged as a pressing concern with the increasing prevalence and accessibility of conversational LLMs. Adversarial users often exploit these models through carefully engineered prompts to elicit restricted or sensitive outputs, a strategy widely referred to as jailbreaking. While numerous defense mechanisms have been proposed, attackers continuously develop novel prompting techniques, and no existing model can be considered fully resistant. In this study, we investigate the jailbreak phenomenon by examining the internal representations of LLMs, with a focus on how hidden layers respond to jailbreak versus benign prompts. Specifically, we analyze the open-source LLM GPT-J and the state-space model Mamba2, presenting preliminary findings that highlight distinct layer-wise behaviors. Our results suggest promising directions for further research on leveraging internal model dynamics for robust jailbreak detection and defense.

As large language models (LLMs) are deployed globally, creating pluralistic systems that can accommodate the diverse preferences and values of users worldwide becomes essential. We introduce EVALUESTEER, a benchmark to measure LLMs' and reward models' (RMs) steerability towards users' value and stylistic preference profiles grounded in psychology and human-LLM interaction literature. To address the gap in existing datasets that do not support controlled evaluations of RM steering, we synthetically generated 165,888 preference pairs -- systematically varying pairs along 4 value dimensions (traditional, secular-rational, survival, and self-expression) and 4 style dimensions (verbosity, readability, confidence, and warmth). We use EVALUESTEER to evaluate whether, given a user profile and a pair of candidate value-laden and style-laden responses, LLMs and RMs are able to select the output that aligns with the user's preferences. We evaluate six open-source and proprietary LLMs and RMs under eleven systematic prompting conditions and six preference comparison scenarios. Notably, our results show that, when given the user's full profile of values and stylistic preferences, the best models achieve <75% accuracy at choosing the correct response, in contrast to >99% accuracy when only relevant style and value preferences are provided. EVALUESTEER thus highlights the limitations of current RMs at identifying and adapting to relevant user profile information, and provides a challenging testbed for developing RMs that can be steered towards diverse human values and preferences.

The study of emergent behaviors in large language model (LLM)-driven multi-agent systems is a critical research challenge, yet progress is limited by a lack of principled methodologies for controlled experimentation. To address this, we introduce Shachi, a formal methodology and modular framework that decomposes an agent's policy into core cognitive components: Configuration for intrinsic traits, Memory for contextual persistence, and Tools for expanded capabilities, all orchestrated by an LLM reasoning engine. This principled architecture moves beyond brittle, ad-hoc agent designs and enables the systematic analysis of how specific architectural choices influence collective behavior. We validate our methodology on a comprehensive 10-task benchmark and demonstrate its power through novel scientific inquiries. Critically, we establish the external validity of our approach by modeling a real-world U.S. tariff shock, showing that agent behaviors align with observed market reactions only when their cognitive architecture is appropriately configured with memory and tools. Our work provides a rigorous, open-source foundation for building and evaluating LLM agents, aimed at fostering more cumulative and scientifically grounded research.

Large language models (LLMs) are increasingly deployed as evaluators of text quality, yet the validity of their judgments remains underexplored. This study investigates systematic bias in self- and cross-model evaluations across three prominent LLMs: ChatGPT, Gemini, and Claude. We designed a controlled experiment in which blog posts authored by each model were evaluated by all three models under four labeling conditions: no attribution, true attribution, and two false-attribution scenarios. Evaluations employed both holistic preference voting and granular quality ratings across three dimensions Coherence, Informativeness, and Conciseness with all scores normalized to percentages for direct comparison. Our findings reveal pronounced asymmetries in model judgments: the "Claude" label consistently elevated scores regardless of actual authorship, while the "Gemini" label systematically depressed them. False attribution frequently reversed preference rankings, producing shifts of up to 50 percentage points in voting outcomes and up to 12 percentage points in quality ratings. Notably, Gemini exhibited severe self-deprecation under true labels, while Claude demonstrated intensified self-preference. These results demonstrate that perceived model identity can substantially distort both high-level judgments and fine-grained quality assessments, independent of content quality. Our findings challenge the reliability of LLM-as-judge paradigms and underscore the critical need for blind evaluation protocols and diverse multi-model validation frameworks to ensure fairness and validity in automated text evaluation and LLM benchmarking.

Causal inference in spatial domains faces two intertwined challenges: (1) unmeasured spatial factors, such as weather, air pollution, or mobility, that confound treatment and outcome, and (2) interference from nearby treatments that violate standard no-interference assumptions. While existing methods typically address one by assuming away the other, we show they are deeply connected: interference reveals structure in the latent confounder. Leveraging this insight, we propose the Spatial Deconfounder, a two-stage method that reconstructs a substitute con-founder from local treatment vectors using a conditional variational autoencoder (CV AE) with a spatial prior, then estimates causal effects via a flexible outcome model. We show that this approach enables nonparametric identification of both direct and spillover effects under weak assumptions--without requiring multiple treatment types or a known model of the latent field. Empirically, we extend SpaCE, a benchmark suite for spatial confounding, to include treatment interference, and show that the Spatial Deconfounder consistently improves effect estimation across real-world datasets in environmental health and social science. By turning interference into a multi-cause signal, our framework bridges spatial and deconfounding literatures to advance robust causal inference in structured data. Causal inference in spatial settings is critical for science and policy, from estimating the health effects of pollution to evaluating land use, climate interventions, and the spread of infectious disease. Most data in these domains are observational, since large-scale interventions are typically infeasible or unethical, so robust methodology is needed to draw valid conclusions. Y et observational studies in these settings face two fundamental challenges that standard methods rarely address together: (1) spillover (interference), where the treatment at one site affects outcomes at nearby sites, violating the Stable Unit Treatment V alue Assumption (SUTV A), and (2) spatially structured unobserved confounding, where latent fields such as weather or socioeconomic context jointly drive exposures and outcomes.

Gabriela Cibils is on a mission - to help turn Paraguay into the Silicon Valley of South America. When she was growing up in the landlocked country, nestled between Brazil and Argentina, she says the nation wasn't super tech focused. But it was different for Ms Cibils, as her parents worked in the technology sector. And she was inspired to study in the US, where she got a degree in computing and neuroscience from the University of California, Berkeley. After graduating she spent eight years working in Silicon Valley, near San Francisco, with roles at various American start-ups.

Robby Starbuck speaks in an interview in New York in March. Robby Starbuck speaks in an interview in New York in March. Critics condemn Robby Starbuck, appointed in lawsuit settlement, for'peddling lies and pushing extremism' A prominent anti-DEI campaigner appointed by Meta in August as an adviser on AI bias has spent the weeks since his appointment spreading disinformation about shootings, transgender people, vaccines, crime, and protests. Robby Starbuck, 36, of Nashville, was appointed in August as an adviser by Meta - owner of Facebook, Instagram, WhatsApp, and other tech platforms - in an August lawsuit settlement. Since his appointment, Starbuck has baselessly claimed that individual shooters in the US were motivated by leftist ideology, described faith-based protest groups as communists, and without evidence tied Democratic lawmakers to murders.

Charlotte, a spider-like construction robot developed by Crest Robotics and Earthbuilt Technology, can 3D print a 2,150-square-foot home in just 24 hours.

Drone footage has emerged showing a blaze destroying the historic Bernaga Monastery in Italy. Founded in La Valletta Brianza in 1628, it is located about 30km (19 miles) east of Milan. More than 20 cloistered nuns were evacuated from the scene, according to Italian media reports. Could a Corrie cameo be on the cards for Daniel O'Donnell? Daniel O'Donnell said making a cameo on Coronation Street is on his bucket list.

Tommy Flowers: nothing like the machine he proposed had ever been contemplated. Tommy Flowers: nothing like the machine he proposed had ever been contemplated. Move over, Alan Turing: meet the working-class hero of Bletchley Park you didn't see in the movies The Oxbridge-educated boffin is feted as the codebreaking genius who helped Britain win the war. But should a little-known Post Office engineer named Tommy Flowers be seen as the real father of computing? T his is a story you know, right? It's early in the war and western Europe has fallen. Only the Channel stands between Britain and the fascist yoke; only Atlantic shipping lanes offer hope of the population continuing to be fed, clothed and armed. But hunting "wolf packs" of Nazi U-boats pick off merchant shipping at will, coordinated by radio instructions the Brits can intercept but can't read, thanks to the fiendish Enigma encryption machine.