LLMs

It has become fashionable for AI researchers to pepper their papers with references to psychology: System 1 and 2 thinking, Theory of Mind, memory systems, even empathy. But according to a recent meta-analysis titled “The Incomplete Bridge: How AI Research (Mis)Engages with Psychology”, these references are often little more than conceptual garnish. The authors analyze 88 AI papers from NeurIPS and ACL (2022-2023) that cite psychological concepts. Their verdict is sobering: while 78% use psychology as inspiration, only 6% attempt to empirically validate or challenge psychological theories. Most papers cite psychology in passing — using it as window dressing to make AI behaviors sound more human-like. ...

When you tell an AI, “make the third bar blue,” what does it actually see? If it’s a typical large language model (LLM), it doesn’t really see anything. It parses your instruction, guesses what “third bar” means, and fumbles to write chart code—often missing the mark. ChartM$^3$ (Multimodal, Multi-level, Multi-perspective) changes the game. It challenges AIs to not only read and write code but also visually comprehend what a user points at. With 1,000 human-curated chart editing tasks and 24,000 training examples, this new benchmark sets a higher bar—one that demands both verbal and visual fluency. ...

In the escalating arms race between fraudsters and detection systems, recent advances in Graph-Enhanced LLMs hold enormous promise. But they face a chronic problem: too much information. Take graph-based fraud detection. It’s common to represent users and their actions as nodes and edges on a heterogeneous graph, where each node may contain rich textual data (like reviews) and structured features (like ratings). To classify whether a node (e.g., a user review) is fraudulent, models like GraphGPT or HiGPT transform local neighborhoods into long textual prompts. But here’s the catch: real-world graphs are dense. Even two hops away, the neighborhood can balloon to millions of tokens. ...

When ChatGPT makes up a statistic or misstates a date, it’s annoying. But when a financial assistant claims the wrong interest expense or misattributes a revenue source, it could move markets or mislead clients. This is the stark reality FRED confronts head-on. FRED—short for Financial Retrieval-Enhanced Detection and Editing—is a framework fine-tuned to spot and fix factual errors in financial LLM outputs. Developed by researchers at Pegasi AI, it isn’t just another hallucination detection scheme. It’s an auditor with a domain-specific brain. ...

What happens when your AI co-pilot thinks it’s the pilot? In financial decision-making, autonomy isn’t always a virtue. A striking new study titled “Your AI, Not Your View” reveals that even the most advanced Large Language Models (LLMs) may quietly sabotage your investment strategy — not by hallucinating facts, but by overriding your intent with stubborn preferences baked into their training. Hidden Hands Behind the Recommendations The paper introduces a systematic framework to identify and measure confirmation bias in LLMs used for investment analysis. Instead of just summarizing news or spitting out buy/sell signals, the study asks: what if the model already has a favorite? More specifically: ...

Social simulations are entering their uncanny valley. Fueled by generative agents powered by Large Language Models (LLMs), recent frameworks like Smallville, AgentSociety, and SocioVerse simulate thousands of lifelike agents forming friendships, spreading rumors, and planning parties. But do these simulations reflect real social processes — or merely replay the statistical shadows of the internet? When Simulacra Speak Fluently LLMs have demonstrated striking abilities to mimic human behaviors. GPT-4 has passed Theory-of-Mind (ToM) tests at levels comparable to 6–7 year-olds. In narrative contexts, it can detect sarcasm, understand indirect requests, and generate empathetic replies. But all of this arises not from embodied cognition or real-world goals — it’s just next-token prediction trained on massive corpora. ...

Most tool-augmented LLMs approach math reasoning like they’re wielding a hammer—good for hitting one nail at a time, but ill-equipped when the problem requires a wrench, a compass, and a soldering iron all at once. Enter Multi-TAG, a clever, finetuning-free framework that aggregates the strengths of multiple tools per reasoning step. Think of it as an LLM with a toolbox, not just a single tool. And it doesn’t just work—it wins, posting 6.0% to 7.5% accuracy gains across MATH500, AIME, AMC, and OlympiadBench against top baselines, using both open and closed LLMs. ...

Fine-tuning is the hammer; steering is the scalpel. In an era where models are increasingly opaque and high-stakes, we need tools that guide behavior without overhauling the entire architecture. That’s precisely what GRAINS (Gradient-based Attribution for Inference-Time Steering) delivers: a powerful, interpretable, and modular way to shift the behavior of LLMs and VLMs by leveraging the most fundamental unit of influence—the token. The Problem with Global Steering Traditional inference-time steering approaches often rely on global intervention vectors: a blunt, one-size-fits-all shift in hidden activations derived from paired desirable and undesirable examples. But these methods are insensitive to which specific tokens caused bad behavior. It’s like adjusting a recipe because the dish tastes bad—without checking if the salt or the sugar was at fault. ...

When we talk about parameter-efficient fine-tuning, LoRA (Low-Rank Adaptation) is often celebrated as a silver bullet: cost-effective, memory-efficient, and—many assume—safe. After all, it modifies only a small fraction of model parameters, sideloaded as low-rank matrices, while leaving the massive pretrained model backbone untouched. The prevailing belief has been that such minimal intervention can’t possibly memorize or leak sensitive data. This belief is now decisively debunked by LoRA-Leak, a landmark framework introduced in a new paper by researchers from Tsinghua and HKUST. Their findings are a wake-up call for AI developers and policymakers alike: even LoRA-finetuned models are highly vulnerable to membership inference attacks (MIAs)—and ironically, the very presence of the frozen pretrained model amplifies this leakage risk. ...

Climate science, once defined by hand-tuned code and static diagnostics, is entering a new phase of automation and adaptability. At the forefront is EarthLink, a self-evolving multi-agent AI platform built specifically to support Earth system science. But this isn’t another LLM wrapper for answering climate questions. EarthLink is something deeper: a scientific collaborator that plans experiments, writes code, debugs itself, interprets results, and learns with each use. From Toolkits to Thinking Partners Traditional tools like ESMValTool or ILAMB have standardized climate model evaluation, but they remain brittle and rigid. They require domain-specific programming expertise and offer little flexibility beyond predefined tasks. In contrast, EarthLink introduces a new paradigm: ...