Cognaptus Insights

In an age where TikTok traders and YouTube gurus claim market mastery, a new benchmark dataset asks a deceptively simple question: Can AI tell when someone really believes in their own stock pick? The answer, it turns out, reveals not just a performance gap between finfluencers and index funds, but also a yawning chasm between today’s multimodal AI models and human judgment. Conviction Is More Than a Call to Action The paper “VideoConviction” introduces a unique multimodal benchmark composed of 288 YouTube videos from 22 financial influencers, or “finfluencers,” spanning over six years of market cycles. From these, researchers extracted 687 stock recommendation segments, annotating each with: ...

What makes a large language model (LLM) biased? Is it the instruction tuning data, the randomness of training, or something more deeply embedded? A new paper from Itzhak, Belinkov, and Stanovsky, presented at COLM 2025, delivers a clear verdict: pretraining is the primary source of cognitive biases in LLMs. The implications of this are far-reaching — and perhaps more uncomfortable than many developers would like to admit. The Setup: Two Steps, One Core Question The authors dissected the origins of 32 cognitive biases in LLMs using a controlled two-step causal framework: ...

When it comes to prompting vision-language models, most methods rely on textual descriptions extracted from large language models like GPT. But those descriptions—“fluffy fur, friendly eyes, golden color”—are often verbose, ambiguous, or flat-out unreliable. What if we could skip that noisy middle step entirely? That’s the premise behind DeMul (Description-free Multi-prompt Learning), a new method presented at ICLR 2025 that quietly delivers a major leap in few-shot image classification. Instead of generating descriptions for each class, DeMul directly distills the semantic knowledge of GPT embeddings into learnable prompt vectors. The result is simpler, more robust, and strikingly effective. ...

The search for new superconductors, energy materials, and exotic compounds often begins not in a lab—but in a database. Yet despite decades of digitization, scientific knowledge remains fragmented across millions of papers, scattered ontologies, and uncharted connections. A new study from Los Alamos National Laboratory proposes an AI-driven framework that doesn’t just analyze documents—it predicts the next breakthrough. From Papers to Properties: A Three-Tiered Approach At the heart of this method is a clever ensemble pipeline that combines interpretability with predictive power. The authors start by mapping over 46,000 papers on transition-metal dichalcogenides (TMDs)—a key class of 2D materials—into a matrix of latent topics and material mentions. Then they apply a hierarchical modeling approach: ...

When was the last time a machine truly surprised you—not with a quirky ChatGPT poem or a clever image generation, but with scientific reasoning that evolved on its own? Meet STELLA, an AI agent for biomedical research that doesn’t just solve problems—it gets better at solving them while solving them. The Static Curse of Smart Agents Modern AI agents have shown promise in navigating the labyrinth of biomedical research, where each inquiry might require cross-referencing papers, running custom bioinformatics analyses, or interrogating molecular databases. But the vast majority of these agents suffer from a fatal limitation: they rely on static, pre-installed toolkits and hard-coded logic trees. Like a PhD student who memorized a textbook but never updated it, they can’t adapt to new tasks or new knowledge without human intervention. ...

The best kind of privacy leak is the one you can measure. A recent paper by Huang et al. introduces a deceptively simple but powerful principle—the Entropy-Memorization Law—that allows us to do just that. It claims that the entropy of a text sequence is strongly correlated with how easily it’s memorized by a large language model (LLM). But don’t mistake this for just another alignment paper. This law has concrete implications for how we audit models, design prompts, and build privacy-aware systems. Here’s why it matters. ...

If AI is going to understand people, it first has to understand relationships. But when it comes to parsing character connections from narrative texts — whether news articles, biographies, or novels — even state-of-the-art language models stumble. They hallucinate links, miss cross-sentence cues, and often forget what they’ve just read. Enter SymbolicThought, a hybrid framework that gives LLMs a logic-boosted sidekick: symbolic reasoning. Developed by researchers at King’s College London and CUHK, the system doesn’t just extract character relationships from text; it builds editable graphs, detects logical contradictions, and guides users through verification with a smart, interactive interface. ...

“In the absence of perfect knowledge, how do we still reason causally?” This paper tackles a profound and practical dilemma in causal inference: what if we don’t know the full causal graph? In real-world settings — whether in healthcare, finance, or digital platforms — complete causal diagrams are rare. Practitioners instead rely on causal abstractions: simplified, coarse-grained representations that preserve partial causal knowledge. But this raises a fundamental question: Which causal queries can still be identified under such abstraction? ...

Residual Entanglement: How ResQuNNs Fix Gradient Flow in Quantum Neural Networks In classical deep learning, residual connections revolutionized the training of deep networks. Now, a similar breakthrough is happening in quantum machine learning. The paper “ResQuNNs: Towards Enabling Deep Learning in Quantum Convolution Neural Networks” introduces a method to overcome a fundamental bottleneck in Quantum Convolutional Neural Networks (QuNNs): the inability to train multiple quantum layers due to broken gradient flow. ...

The Meek Shall Compute It For the past five years, discussions about AI progress have centered on a simple formula: more data + more compute = better models. This scaling paradigm has produced marvels like GPT-4 and Gemini—but also entrenched a new aristocracy of compute-rich players. Is this inequality here to stay? According to a provocative new paper from MIT CSAIL, the answer may be: not for long. The authors argue that due to the laws of diminishing returns, the performance gap between state-of-the-art (SOTA) models and smaller, cheaper “meek” models will shrink over time. If true, this reframes the future of AI as one not of centralized supremacy, but of widespread, affordable competence. ...