AI News Archive: May 7, 2026 — Part 21
Sourced from 500+ daily AI sources, scored by relevance.
- Grokking or Glitching? How Low-Precision Drives Slingshot Loss Spikes
Deep neural networks exhibit periodic loss spikes during unregularized long-term training, a phenomenon known as the "Slingshot Mechanism." Existing work usually attributes this to intrinsic optimization dynamics, but its triggering mechanism remains unclear. This paper proves that this phenomenon i...
- Towards Reliable LLM Evaluation: Correcting the Winner's Curse in Adaptive Benchmarking
Adaptive prompt and program search makes LLM evaluation selection-sensitive. Once benchmark items are reused inside tuning, the observed winner's score need not estimate the fresh-data performance of the full tune-then-deploy procedure. We study inference for this procedure-level target under explic...
- Tuning Derivatives for Causal Fairness in Machine Learning
Artificial-intelligence systems are becoming ubiquitous in society, yet their predictions typically inherit biases with respect to protected attributes such as race, gender, or age. Classical fairness notions, most notably Statistical Parity (SP), demand that predictions be independent of the protec...
- CITE: Anytime-Valid Statistical Inference in LLM Self-Consistency
Large language models often improve reasoning by sampling multiple outputs and aggregating their final answers, but precise and efficient control of error levels remains a challenging task. In particular, deciding when to stop sampling remains difficult when the stopping rule is data-dependent and t...
- Ratio-based Loss Functions
Algorithms in machine learning and AI do critically depend on at least three key components: (i) the risk function, which is the expectation of the loss function, (ii) the function space, which is often called the hypothesis space, and (iii) the set of probability measures, which are allowed for the...
- Transformers Provably Implement In-Context Reinforcement Learning with Policy Improvement
We investigate the ability of transformers to perform in-context reinforcement learning (ICRL), where a model must infer and execute learning algorithms from trajectory data without parameter updates. We show that a linear self-attention transformer block can provably implement policy-improvement me...
- Spectral Lens: Activation and Gradient Spectra as Diagnostics of LLM Optimization
Training loss and throughput can hide distinct internal representation in language-model training. To examine these hidden mechanics, we use spectral measurements as practical and operational diagnostics. Using a controlled family of decoder-only models adapted from the modded NanoGPT codebase, we i...
- Neural-Actuarial Longevity Forecasting: Anchoring LSTMs for Explainable Risk Management
Traditional multi-population models, such as the Li-Lee framework, rely on the assumption of mean-reverting country-specific deviations. However, recent data from high-longevity clusters suggest a systemic break in this paradigm. We identify a stationarity paradox where mortality residuals in countr...
- Decoupled PFNs: Identifiable Epistemic-Aleatoric Decomposition via Structured Synthetic Priors
Prior-Fitted Networks (PFNs) amortize Bayesian prediction by meta-learning over a synthetic task prior, but their standard output is a posterior predictive distribution over noisy observations. For sequential decision-making, such as active learning and Bayesian optimization, acquisition should prio...
- Beyond the Independence Assumption: Finite-Sample Guarantees for Deep Q-Learning under $τ$-Mixing
Finite-sample analyses of deep Q-learning typically treat replayed data as independent, even though it is sampled from temporally dependent state-action trajectories. We study the Deep Q-networks (DQN) algorithm under explicit dependence by modelling the minibatches used for updating the network as ...
- The Interplay of Data Structure and Imbalance in the Learning Dynamics of Diffusion Models
Real-world datasets are inherently heterogeneous, yet how per-class structural differences and sampling imbalance shape the training dynamics of diffusion models-and potentially exacerbate disparities-remains poorly understood. While models typically transition from an initial phase of generalizatio...
- Topological Signatures of Grokking
We study the grokking phenomenon through the lens of topology. Using persistent homology on point clouds derived from the embedding matrices of a range of models trained on modular arithmetic with varying primes, we identify a clear and consistent topological signature of grokking: a sharp increase ...
- End-to-End Identifiable and Consistent Recurrent Switching Dynamical Systems
Learning identifiable representations in deep generative models remains a fundamental challenge, particularly for sequential data with regime-switching dynamics. Existing approaches establish identifiability under restrictive assumptions, such as stationarity or limited emission models, and typicall...
- Attributions All the Way Down? The Metagame of Interpretability
We introduce the metagame, a conceptual framework for quantifying second-order interaction effects of model explanations. For any first-order attribution $φ(f)$ explaining a model $f$, we measure the directional influence of feature $j$ on the attribution of feature $i$, denoted as meta-attribution ...
- Expressivity of Bi-Lipschitz Normalizing Flows: A Score-Based Diffusion Perspective
Many normalizing flow architectures impose regularity constraints, yet their distributional approximation properties are not fully characterized. We study the expressivity of bi-Lipschitz normalizing flows through the lens of score-based diffusion models. For the probability flow ODE of a variance-p...
- TabCF: Distributional Control Function Estimation with Tabular Foundation Models
Instrumental variable (IV) and control function (CF) methods are powerful tools for causal effect estimation in the presence of unmeasured confounding, yet most existing approaches target only mean effects and/or demand substantial fitting and tuning effort. In this paper, we introduce a simple meth...
- Temporal Functional Circuits: From Spline Plots to Faithful Explanations in KAN Forecasting
Unlike MLPs, Kolmogorov-Arnold Networks (KANs) expose explicit learnable edge functions on every connection, enabling mechanistic explanation in time-series forecasting. This paper introduces Temporal Functional Circuits, a framework that transforms KAN edge functions from latent visualizations into...
- Spherical Flows for Sampling Categorical Data
We study the problem of learning generative models for discrete sequences in a continuous embedding space. Whereas prior approaches typically operate in Euclidean space or on the probability simplex, we instead work on the sphere $\mathbb S^{d-1}$. There the von Mises-Fisher (vMF) distribution induc...
- Variational Smoothing and Inference for SDEs from Sparse Data with Dynamic Neural Flows
Stochastic differential equations (SDEs) provide a flexible framework for modeling temporal dynamics in partially observed systems. A central task is to calibrate such models from data, which requires inferring latent trajectories and parameters from sparse, noisy observations. Classical smoothing m...
- In-Context Positive-Unlabeled Learning
Positive-unlabeled (PU) learning addresses binary classification when only a set of labeled positives is available alongside a pool of unlabeled samples drawn from a mixture of positives and negatives. Existing PU methods typically require dataset-specific training or iterative optimization, which l...
- FlowMarket
A social network of AI agents generating B2B deals
- Claude Agents for Financial Services
Finance agent templates for pitches, KYC, and closing books
- Genrate.ai
The military-grade recon machine for revenue teams.
- Luma Uni 1.1 API
A reasoning model that interprets intent before it generates
- Basedash MCP server
Your data analyst, in every AI tool you already use
- SLED AI
Public-sector revenue engine for B2B companies
- Phrony
Ship AI agents without the operational burden
- Bagel AI
AI product intelligence for product and GTM teams
- DevPass by LLM Gateway
One key to access every coding model in 3 flat prices
- RAKOR
Custom CRM and AI automation for businesses
- Forge
A complete React toolkit made for AI
- Contextual Moderation for Chat
AI-powered moderation for safer chat experiences
- Hachigo
Turn repetitive AI tasks into apps
- Safebooks AI
AI agents for finance operations
- SaolaAI
Autonomous quality for engineering teams
- WINN.AI
Real-time sales copilot for customer calls
- TravelMaxing | AI powered travel agency
The quality of travel agencies without the heavy fees
- Inkbox
Give your AI agents email, phone and an internet address
- tilde.run
Serverless sanbox for agents, with a versioned filesystem.
- Hyperflow
AI-Powered Intelligence For Enterprise On-Chain Data
- AiDesigns — AI creative workspace
Chat agent and canvas for top image, video & music models.
- Grepture
The gateway for every AI call your app makes.
- Natural Language Autoencoders Produce Unsupervised Explanations of LLM Activations
Abstract We introduce Natural Language Autoencoders (NLAs), an unsupervised method for generating natural language explanations of LLM activations. An NLA consists of two LLM modules: an activation verbalizer (AV) that maps an activation to a text description and an activation reconstructor (AR) that maps the description back to an activation. We jointly train the AV and AR with reinforcement learning to reconstruct residual stream activations. Although we optimize for activation reconstruction, the resulting NLA explanations read as plausible interpretations of model internals that, according to our quantitative evaluations, grow more informative over training. We apply NLAs to model auditing. During our pre-deployment audit of Claude Opus 4.6, NLAs helped diagnose safety-relevant behaviors and surfaced unverbalized evaluation awareness—cases where Claude believed, but did not say, that it was being evaluated. We present these audit findings as case studies and corroborate them using
- Mechanistic estimation for wide random MLPs
This post covers joint work with Wilson Wu, George Robinson, Mike Winer, Victor Lecomte and Paul Christiano. Thanks to Geoffrey Irving and Jess Riedel for comments on the post. In ARC's latest paper, we study the following problem: given a randomly initialized multilayer perceptron (MLP), produce an estimate for the expected output of the model under Gaussian input. The usual approach to this problem is to sample many possible inputs, run them all through the model, and take the average. Instead, we produce an estimate "mechanistically", without running the model even once. For wide models, our approach produces more accurate estimates, both in theory and in practice. Paper: Estimating the expected output of wide random MLPs more efficiently than sampling Code: mlp_cumulant_propagation GitHub repo We are excited about this result as an early step towards our goal of producing mechanistic estimates that outperform random sampling for any trained neural network. Drawing an analogy betwee
- Sustaining Cooperation in Populations Guided by AI: A Folk Theorem for LLMs
Large language models (LLMs) are increasingly used to provide instructions to many agents who interact with one another. Such shared reliance couples agents who appear to act independently: they may in fact be guided by a common model. This coupling can change the prospects for cooperation among age...
- AgenticPrecoding: LLM-Empowered Multi-Agent System for Precoding Optimization
Precoding is a key technique for interference management and performance improvement in multi-antenna wireless systems. However, existing precoding methods are typically developed for specific system models, objectives, and constraint sets, which limits their adaptability to the heterogeneous and ev...
- Independent Learning of Nash Equilibria in Partially Observable Markov Potential Games with Decoupled Dynamics
We study Nash equilibrium learning in partially observable Markov games (POMGs), a multi-agent reinforcement learning framework in which agents cannot fully observe the underlying state. Prior work in this setting relies on centralization or information sharing, and suffers from sample and computati...
- From Agent Loops to Deterministic Graphs: Execution Lineage for Reproducible AI-Native Work
Large language model systems are increasingly deployed as agentic workflows that interleave reasoning, tool use, memory, and iterative refinement. These systems are effective at producing answers, but they often rely on implicit conversational state, making it difficult to preserve stable work produ...
- Active Learning for Communication Structure Optimization in LLM-Based Multi-Agent Systems
Optimizing the communication structure of large language model based multi-agent systems (LLM-MAS) has been shown to improve downstream performance and reduce token usage. Existing methods typically rely on randomly sampled training tasks. However, tasks may differ substantially in difficulty and do...
- Retrieval-Conditioned Topology Selection with Provable Budget Conservation for Multi-Agent Code Generation
Multi-agent LLM systems for code generation face a fundamental routing problem: the optimal orchestration topology depends on the structural complexity of the code under modification, yet existing systems select topologies without consulting the codebase. We present Retrieval-Guided Adaptive Orchest...