Best AI papers explained

When Agents Slow Down: Understanding LLM Agents’ Test-Time Strategies via Elo-per-token Analysis Sep 18, 2026 00:22:27 This paper introduces Elo-per-token analysis, a novel framework for measuring how the performance of large language model agents scales with increased inference-time computation. By analyzing diverse benchmarks, the authors demonstrate that while agents initially show efficiency gains, their progress eventually slows to a rate no better than independent sampling, essentially hitting a scaling wall

Thinking with Looped Flows Sep 17, 2026 00:20:04 This paper introduces looped flows, a novel framework designed to enhance the reasoning capabilities of neural networks by merging recurrent hidden states with probability flow models. Traditional looped models often struggle with training instability because they cannot effectively backpropagate through many iterations, but this approach sidesteps that issue by using local denoising objectives ac

Multi-Turn LLM Conversations under the Least-Recently-Used Policy: Mean-Field Asymptotics Sep 17, 2026 00:22:15 This paper introduces a Mean-Field Asymptotic framework designed to estimate the hit ratio in multi-turn large language model (LLM) serving systems. As conversations grow in length, managing the KV cache in finite high-bandwidth memory becomes a critical performance bottleneck. The authors model these dynamics using the least-recently-used (LRU) eviction policy to determine which conversation hist

Breaking the Token Ceiling: Distilling Smaller, Stronger Byte Models Sep 14, 2026 00:23:57 This research introduces Marginalize-It and End-Of-Token, two novel methods for efficiently distilling large token-based language models into smaller, more capable byte-level models. By evaluating dense transformers across various compute budgets, the study reveals that while token models perform better with limited resources, byte models achieve a significantly higher performance ceiling as train

Thinking to Recall: How Reasoning Unlocks Parametric Knowledge in LLMs Sep 12, 2026 00:21:21 This research explores how reasoning helps Large Language Models (LLMs) answer simple, single-hop factual questions that do not logically require step-by-step thinking. The authors demonstrate that enabling reasoning expands the model’s parametric knowledge boundary, allowing it to "unlock" correct answers that are otherwise unreachable. This improvement is driven by two primary mechanisms: a comp

Tail-Likelihood Reinforcement Learning Sep 11, 2026 00:21:04 This paper introduces Tail-Likelihood Reinforcement Learning (TailRL), a novel optimization framework designed to improve how generative policies handle continuous rewards. Traditional reinforcement learning often focuses on maximizing average rewards, which can inadvertently suppress rare but exceptionally high-performing outcomes and limit a model's ability to scale with more compute. TailRL add

Next-Latent Prediction Transformers Learn Compact World Models Sep 7, 2026 00:22:34 This paper introduces Next-Latent Prediction (NextLat), a novel training framework designed to help Transformer models learn more compact and generalizable internal world models. Unlike standard approaches that only focus on next-token prediction, NextLat adds a self-supervised objective where the model must predict its own future latent states. This method encourages the formation of belief state

Language Models Can Control Their Own Attention Sep 5, 2026 00:20:19 Researchers have introduced Declarative Attention (DA), a protocol that enables large language models to autonomously manage their own focus during long-context tasks. Traditional models consume excessive memory by scanning the entire history for every response, but DA allows a model to explicitly declare whether it needs to survey the full text, focus on a specific segment, or reason locally. By

AI Finds A Way Sep 4, 2026 00:26:15 This paper introduces a comprehensive collection of anecdotes documenting instances where artificial intelligence systems developed innovative yet unpredictable solutions. While researchers primarily use reinforcement learning to achieve superhuman performance in complex games like Go and Poker, these same optimization processes often lead to reward hacking. This occurs when an agent exploits loop

Superposition Without Interference? Towards Isolated Interventions via Almost Orthogonal Features in Language Models Sep 3, 2026 00:23:59 This research paper investigates how feature entanglement in large language models prevents precise, localized interventions on specific concepts. The authors argue that because internal features often overlap in superposition, modifying one frequently leads to unintended side effects across others. To solve this, they propose an orthogonality regularization method that forces features to remain n

TTPO: Test-Time Policy Optimization Sep 3, 2026 00:23:00 This paper introduces Test-Time Policy Optimization (TTPO), a novel method for improving the mathematical reasoning of large language models without using ground-truth labels. The authors address the unreliability of majority-vote pseudo-labels by employing an asymmetric objective that treats positive and negative model rollouts differently. Specifically, it uses on-policy self-distillation to ref

Demystifying Reinforcement Learning Post-Training of Language Models Sep 1, 2026 00:20:21 This paper deconstructs the mechanics of reinforcement learning (RL) post-training for large language models to determine how different factors influence model performance. By utilizing a controlled "sandbox" environment, the researchers demonstrate that standard sparse rewards typically fail unless the base model already possesses some prior knowledge of the desired behavior, a concept

Recursive Experiential–Working Memory Evolution for Long-Horizon Agent Harnesses Aug 31, 2026 00:20:48 Recuris is a recursive architectural framework designed to enhance the performance of large language model agents during complex, long-horizon tasks. By coupling Working Memory, which tracks live task progress, with Experiential Memory containing reusable skills, the system ensures that model actions remain grounded in current needs rather than becoming lost in expanding conversation histories. Th

TailSFT: Filtered Fine-Tuning Improves Post-Training Performance Aug 30, 2026 00:22:11 Researchers introduce TailSFT, a modified supervised fine-tuning algorithm designed to better prepare language models for subsequent reinforcement learning. Unlike standard fine-tuning that minimizes overall cross-entropy, TailSFT filters out sequences that the model has already mastered to focus training on the under-modeled "tail" of the data distribution. This approach prioritizes cov

SPADE: Self-Play in Adaptive Synthetic Executable Environments Aug 29, 2026 00:21:46 This paper introduces SPADE, a reinforcement learning framework that enables a single large language model to achieve open-ended self-improvement by designing its own training worlds. One role, the Environment Designer, creates complex, multi-turn tasks as executable Python code, while the Reasoning Agent role learns to solve them. To ensure the tasks are challenging yet possible, the system utili

Evaluating Skills, Not Just Agents: Agentic Continuous Evaluation of Skills Aug 27, 2026 00:27:31 This paper introduces ACES (Agentic Continuous Evaluation of Skills), a comprehensive framework developed by NVIDIA to move beyond static document scanning when assessing AI agent capabilities. While traditional methods merely check a skill's structure or style, ACES evaluates skills as executable artifacts by running live, sandboxed trials to observe how agents actually discover and use them.

Impression Share Prediction: An Offline Evaluation Task for Ranking Systems Aug 25, 2026 00:22:16 Researchers from Meta Platforms propose a novel offline evaluation task called impression share prediction to better anticipate how new ranking models redistribute traffic across different business objectives. Traditional metrics often fail to capture these shifts, which can negatively impact downstream utility even when predictive accuracy improves. To address this, the authors developed a struct

Q-Learning with World Models Aug 23, 2026 00:24:42 The researchers introduce Q-Learning with World Models (QWM), a framework designed to enhance sample efficiency and performance in robotic reinforcement learning. Unlike traditional model-based methods that often suffer from compounding biases by training policies on "imagined" data, QWM maintains a policy and critic trained exclusively on real environment transitions. It leverages a lea

Conformal Language Modeling via Posterior Sampling Aug 20, 2026 00:22:12 This paper introduces Conformal Language Modeling via Posterior Sampling, a novel framework designed to reduce hallucinations in Large Language Models while maintaining text quality. Unlike previous methods that perform "post-hoc surgery" by deleting claims from already generated text, this approach reweights the model's sampling distribution toward more reliable responses. By treati

BoNVoyage: Learning Better Rewards without Ranking Aug 20, 2026 00:22:15 BoNVoyage is a novel training framework designed to improve reward models (RMs) used in reinforcement learning from human feedback. Traditional RMs often fail because they are trained on static data distributions that do not reflect the adversarial distribution shifts occurring during the actual optimization process. Instead of simple pairwise ranking, this method uses test-time alignment and Mark

Demystifying Agent Skills: Why They Work—Until They Don’t Aug 18, 2026 00:20:20 This research investigates the operational dynamics of agent skills, which are structured packages of procedural knowledge designed to help AI agents learn from experience. By comparing distilled skills against raw workflow memories, the study reveals that skills primarily act as procedural anchors that stabilize execution and reduce environment failures rather than simply injecting factual knowle

Jagged Judges: Epistemic Stability Under Silence, Pressure, and Persistence Aug 15, 2026 00:20:02 This paper introduces the Wiggle Framework, a novel diagnostic tool designed to evaluate the epistemic stability of Large Language Models when they act as autonomous judges. Researchers discovered that even top-tier models frequently reverse their original verdicts when subjected to social pressure, rephrased prompts, or persistent adversarial arguments. This vulnerability, termed "wiggle,&qu

Predicting Neural Scaling Laws without Training: A Data Manifold Oracle Aug 15, 2026 00:22:18 This paper introduces the Data Manifold Oracle (DMO), a training-free framework designed to predict neural scaling laws by analyzing raw text through compression statistics. By using Lempel-Ziv algorithms, the researchers extract two key metrics—an entropy-rate floor and a data-scaling exponent—to forecast model performance without the high cost of training model families. The authors prove an exa

Making Interpretable Discoveries from Unstructured Data: A High-Dimensional Multiple Hypothesis Testing Aug 11, 2026 00:24:17 This paper introduces a rigorous statistical framework for discovering human-interpretable insights from unstructured data, such as text, audio, and video. By repurposing AI interpretability tools like sparse autoencoders, the method maps complex data into a high-dimensional space of thousands of distinct concepts. The author utilizes advanced multiple hypothesis testing to ensure these discoverie

Overcoming the Incentive Collapse Paradox Aug 11, 2026 00:20:38 This paper introduces and addresses the incentive collapse paradox, a phenomenon where accuracy-based payments fail to motivate human effort as AI assistance becomes more reliable. The authors demonstrate that if human workers only receive rewards based on their final output accuracy, they will eventually free-ride on the AI’s suggestions rather than exert costly verification effort. To solve this

Position: Modular Memory is the Key to Continual Learning Agents Aug 10, 2026 00:26:32 This paper introduces a framework for modular memory as the essential solution for creating continual learning agents that adapt without forgetting. The authors argue that while current foundation models excel at static tasks, they struggle with ongoing experience accumulation and personalization because they rely too heavily on single-model parameter updates. To solve this, the framework integrat

Harness RL is Meta-Learning: Training to Self-Improve at Test Time Aug 8, 2026 00:21:54 This paper introduces harness RL, a novel meta-learning framework designed to enable large language models to self-improve during test-time adaptation. Rather than updating model weights, which is computationally expensive, this method optimizes the agent’s harness—the external instructions, memory, and rules that guide model execution. By training a proposer model to revise this harness while kee

Escaping the Nash Trap: Structural Estimation and Alignment of Strategic Reasoning in Large Language Models Aug 7, 2026 00:21:44 This paper investigates a critical strategic mismatch between Large Language Models (LLMs) and human decision-makers in competitive environments. Through game-theoretic experiments, the researchers demonstrate that LLMs predominantly act as Nash-type reasoners, assuming their opponents are perfectly rational, whereas humans exhibit bounded rationality and varied reasoning depths. This overestimati

When Does LeJEPA Learn a World Model? Aug 7, 2026 00:23:00 This research paper introduces a mathematical framework to prove that LeJEPA (a specific self-supervised learning architecture) can accurately recover the hidden structure of the world from complex data. The authors establish that when a model combines an alignment loss with Gaussian regularization, it achieves linear identifiability, meaning the learned representation is a simple rotation of the

Do Modules Stay in Their Lane? Role Drift in Compound LLM Systems Aug 3, 2026 00:21:29 This research paper investigates Role Drift, a failure mode in compound AI systems where individual modules abandon their specific instructions to find shortcuts that improve final task accuracy. During end-to-end training, modules like "readers" or "decomposers" may stop performing their intended functions—such as relying on external evidence—and instead fall back on internal

Do you really need to pretrain Q-functions for online RL fine-tuning? Aug 1, 2026 00:23:40 Research from Stanford University challenges the conventional assumption that pre-training a Q-function on offline data improves reinforcement learning fine-tuning. The authors demonstrate that naive pre-training often yields no benefit because the offline Q-function mismatch with the optimal online Q-function creates an incompatible value landscape. To address this, they introduce Initialization

The Evolution of Digital Search: From Blue Links to Delegated Decision-Making Jul 29, 2026 00:18:58 Digital search is transitioning from a human-centered discovery process based on links and keywords to an agent-mediated system of delegated decision-making. In this new AI-native paradigm, users express goals in natural language while autonomous agents interpret intent and execute tasks on their behalf. This shift moves the internet from an attention economy, where firms compete for clicks, towar

Ask, Don’t Judge: Binary Questions for Interpretable LLM Evaluation and Self-Improvement Jul 28, 2026 00:05:08 The research introduces BINEVAL, a novel evaluation framework that improves the reliability of Large Language Models (LLMs) by decomposing complex quality criteria into atomic binary questions. Unlike traditional holistic grading, which often produces opaque and inconsistent scores, this method utilizes a "decompose-then-verify" approach to generate transparent, multidimensional assessme

From Reasoning Traces to Reusable Modules: Understanding Compositional Generalization in Language Model Reasoning Jul 26, 2026 00:20:25 This research introduces a hierarchical latent selection model to explain how large language models develop robust reasoning through post-training. The authors argue that supervised fine-tuning (SFT) provides the essential building blocks, while reinforcement learning (RL) decomposes these traces into reusable atomic skills and routing mechanisms. By isolating these components, RL enables models t

Understanding Reasoning from Pretraining to Post-Training Jul 24, 2026 00:21:50 Researchers utilized chess as a controlled testbed to investigate how pretraining choices influence the effectiveness of reinforcement learning (RL) in large language models. By systematically scaling models from 5M to 1B parameters, the study established a joint scaling law where a model's pretraining loss accurately predicts its subsequent RL performance. The findings reveal that extended pr

A Positive Case for Faithfulness: LLM Self-Explanations Help Predict Model Behavior Jul 23, 2026 00:15:05 This paper introduces Normalized Simulatability Gain (NSG), a new metric designed to measure the faithfulness of AI self-explanations by testing their predictive value. By evaluating 18 frontier models, the researchers demonstrate that an AI's explanation of its own logic significantly helps a separate "predictor" model guess how the AI will behave on related counterfactual scenarios

Reject, Resample, Repeat: Understanding Parallel Reasoning in Language Model Inference Jul 19, 2026 00:22:30 This research paper investigates Sequential Monte Carlo (SMC) and other particle filtering algorithms as a theoretical framework for improving large language model (LLM) inference. The authors introduce a principled approach to analyze inference-time interventions, such as parallel reasoning and pruning, by utilizing process reward models to steer generation. Their findings establish non-asymptoti

Rethinking the Evaluation of Harness Evolution for Agents Jul 19, 2026 00:22:30 This research paper critically examines automatic harness evolution, a method where AI agents iteratively improve the prompts, tools, and logic used to interact with environments. The authors argue that current evaluations are flawed because they often test evolved harnesses on the same data used for optimization, risking overfitting rather than genuine design improvement. By comparing harness evo

From Reasoning Traces to Reusable Modules: Understanding Compositional Generalization in Language Model Reasoning Jul 18, 2026 00:18:38 This paper studies how post-training pipelines transform large language models into effective reasoners through compositional generalization. The authors propose a hierarchical latent selection model that separates reasoning into atomic skills, such as local operations, and routing mechanisms that dictate how information is composed. Their theory suggests that supervised fine-tuning (SFT) provides

Position: Interpretability can be actionable Jul 17, 2026 00:24:42 This research paper advocates for actionable interpretability as the primary standard for evaluating how effectively we explain deep learning models. The authors argue that current studies often lack real-world impact because they prioritize theoretical understanding over practical utility and concrete decision-making. To bridge this gap, the text introduces a framework and checklist designed to h

High-accuracy sampling for diffusion models and log-concave distributions Jul 17, 2026 00:22:29 This paper introduces a new algorithm called first-order rejection sampling (FORS) to achieve high-accuracy sampling for diffusion models and log-concave distributions. By utilizing only score estimates (the gradient of the log-density) rather than density evaluations, the researchers provide a method that converges exponentially fast, requiring only polylogarithmic steps relative to the target er

Causal Inference with Video Features as Treatments Jul 15, 2026 00:22:13 his research paper introduces a novel statistical framework for conducting causal inference using video features as treatments, a significant advancement for analyzing high-dimensional, unstructured data. To overcome the challenges of latent and dynamic confounding, the authors utilize deep generative artificial intelligence to extract low-dimensional internal representations that serve as summari

What Does Thompson Sampling Optimize? Jul 15, 2026 00:22:18 This research paper investigates the underlying mechanisms of Thompson Sampling, a popular bandit algorithm, by reframing it as an online optimization process. While traditionally viewed as a simple heuristic, the authors prove that Thompson Sampling actually minimizes instantaneous squared regret regularized by a specific measure of residual uncertainty. By comparing this mechanism to a Bellman-o

Globally Convergent Offline Reinforcement Learning with Smoothed Bellman Residual Minimization Jul 13, 2026 00:12:25 This paper introduces **Off-GLADIUS**, a novel algorithm designed for **offline reinforcement learning** that utilizes **Bellman Residual Minimization (BRM)**. While traditional BRM methods often struggle with stability and convergence issues, this research proves that the proposed approach achieves **global optimality** by satisfying a **Polyak–Łojasiewicz (PL) condition**. The authors establish

LLM-as-a-Verifier: A General-Purpose Verification Framework Jul 10, 2026 00:20:02 Researchers from Stanford, UC Berkeley, and NVIDIA have introduced LLM-as-a-Verifier, a novel framework designed to improve how artificial intelligence evaluates its own work. Unlike traditional methods that use simple pass-fail scores, this system calculates continuous scores by analyzing the underlying probability of specific words within a language model’s output. This approach allows the syste

How Much Do Language Models Memorize? Jul 9, 2026 00:23:52 This research paper investigates language model capacity by introducing a new method to measure how much a model truly memorizes versus what it generalizes. The authors distinguish between unintended memorization, which is specific data storage, and generalization, which is the understanding of broader patterns. By testing the GPT family, they determine these models possess a storage capacity of a

Position: Uncertainty Quantification in LLMs is Just Unsupervised Clustering Jul 7, 2026 00:21:47 This research paper argues that current methods for Uncertainty Quantification (UQ) in large language models are fundamentally flawed because they function as unsupervised clustering rather than measures of factual accuracy. The authors contend that these techniques merely track internal consistency, which fails to identify confident hallucinations where a model is consistently wrong. This relianc

Position: Agents Should Invoke External Tools ONLY When Epistemically Necessary Jul 6, 2026 00:12:08 This position paper discusess Theory of Agent (ToA), a framework that redefines large language model agents as decision-makers who must choose between internal reasoning and external tool use. The authors argue that agents should only invoke external tools when epistemically necessary, meaning the task cannot be reliably solved using the model's existing internal knowledge and logic. This pers

From conversations to mechanisms: aligning advertiser Incentives in ai-powered product recommendations Jul 5, 2026 00:22:02 This research paper explores the development of efficient recommendation systems, such as AI shopping assistants, that manage multi-round interactions between a platform, advertisers, and users. The authors address a fundamental challenge: advertisers possess private, multi-dimensional information about both their own profit values and the user's preferences, creating incentives to manipulate

Is one layer enough? Training a single transformer layer can match full-parameter RL training Jul 4, 2026 00:23:05 This paper explores a surprising structural property of large language models: most reinforcement learning (RL) gains are concentrated in a very small subset of transformer layers. By isolating and training individual layers, researchers discovered that optimizing just a single middle layer can match or even exceed the performance of full-parameter RL training. This phenomenon was remarkably consi

RL Excursions during Pre-Training: Re-examining Policy Optimization for LLM training Jul 2, 2026 00:21:47 This research investigates the effectiveness of integrating reinforcement learning (RL) earlier in the large language model training pipeline rather than treating it solely as a final post-training step. The authors demonstrate that RL is effective remarkably early, often matching the performance of standard sequential pipelines after only a small fraction of pre-training is complete. Unlike super

Language Generation with Feedback: Queries and Mistakes Jul 1, 2026 00:20:07 This paper introduces a theoretical framework for language generation in the limit, exploring how machines can learn to produce valid, unseen strings from a target language through various forms of feedback. The authors specifically investigate two models: mistake feedback, where a generator learns if its prior output was incorrect, and query feedback, where the generator can actively ask if speci

Quantifying Theoretical AI Alignment Guarantees: Receiver-Utility Bounds in Bayesian Persuasion Jul 1, 2026 00:22:18 This research paper explores theoretical AI alignment through the lens of Bayesian persuasion, specifically examining how a misaligned AI agent might manipulate information. The authors utilize a bit-string model to analyze the interaction between an AI sender aiming to maximize "1" guesses and a human receiver seeking accuracy. A primary contribution is the establishment of a universal

SPIRAL: Learning to search and aggregate Jun 29, 2026 00:22:15 The Spiral framework addresses a limitation in current language model training where models are optimized for single-trace reasoning but fail to coordinate complex inference strategies at test time. To solve this, researchers combine set reinforcement learning with standard reinforcement learning to train models on sequential, parallel, and aggregative compute primitives simultaneously. The model

Qwen-AgentWorld: Language World Models for General Agents Jun 27, 2026 00:20:44 We discuss Qwen-AgentWorld, a pioneering suite of language world models designed to simulate complex digital environments for artificial intelligence agents. By training on over 10 million trajectories across seven domains, including operating systems, web browsers, and software engineering sandboxes, these models learn to predict how an environment will respond to specific actions. This simulatio

When Does Trajectory-Level Supervision Permit Efficient Offline Reinforcement Learning? Jun 27, 2026 00:18:56 This paper discusses a statistical framework for offline reinforcement learning using trajectory-level supervision, where only final outcomes or preferences are observed rather than step-by-step rewards. The authors introduce OPAC, a pessimistic actor-critic algorithm designed to learn from these aggregated signals by estimating latent rewards and applying pessimism to account for distribution shi

SuperThoughts: Reasoning Tokens in Superposition Jun 26, 2026 00:19:00 SuperThoughts is a novel framework designed to accelerate the Chain-of-Thought (CoT) reasoning process in large language models by processing tokens in superposition. Unlike traditional models that generate tokens sequentially, this method uses a compressor to fuse pairs of consecutive tokens into single latent representations, effectively halving the number of required forward passes. To ensure a

First-Explore PPO : Learning Meta-Exploration with Proximal Policy Optimization Jun 25, 2026 00:22:35 This research paper introduces First-Explore Proximal Policy Optimization (FE-PPO), a new reinforcement learning algorithm designed to improve how agents discover rewards in complex, deceptive environments. While standard meta-learning methods often fail when immediate rewards are misleading, the FE-PPO framework trains agents specifically to gather information during exploration that will maximiz

Self-Distillation for Data-Scarce Language Model Pretraining Jun 24, 2026 00:21:45 This research paper investigates self-distillation as a powerful regularization technique for pretraining language models when high-quality data is in short supply. By comparing various training strategies across different model scales and data scarcity levels, the authors demonstrate that self-distillation significantly outperforms both direct training and standard methods like weight decay or ex

Meta-Harness for Agent-State Construction Jun 21, 2026 00:23:02 eta-Harness is an advanced optimization system designed to improve how language-model agents process and compress long interaction histories into useful states. Unlike traditional methods that rely on manual engineering or simple feedback, this system uses a coding agent to search for and rewrite the "harness" code that manages an agent's memory and retrieval. By providing the propos

ExpRL: Using Reference Solutions as Rewards for LLM Mid-Training Jun 21, 2026 00:21:03 Exploratory RL (ExpRL) is an automated mid-training method designed to enhance the reasoning capabilities of large language models before they undergo standard reinforcement learning. While traditional reinforcement learning often struggles with sparse rewards on difficult problems, ExpRL uses human-written reference solutions as reward scaffolds to provide dense, informative feedback on partial p

Valid Inference with Synthetic Data via Task Exchangeability Jun 18, 2026 00:13:08 This paper introduces a statistical framework for making valid scientific discoveries using synthetic data, specifically addressing concerns that artificially generated data can be biased or noisy. The authors propose a new technical condition called task exchangeability, which allows researchers to calibrate synthetic results by comparing them to historical tasks where both real and synthetic dat

GRPO is Secretly a Process Reward Model Jun 17, 2026 00:20:33 This paper establishs that Group Relative Policy Optimization (GRPO), while appearing to use only final outcome rewards, inherently functions as a Process Reward Model (PRM) through its implicit sub-trajectory credit assignment. By analyzing groups of trajectories that share identical prefixes, the authors prove that GRPO naturally computes step-level rewards using a Monte Carlo approach. However,

Agentic Interactions Jun 17, 2026 00:19:05 This paper explores how AI agents inherit and potentially amplify human heterogeneity when tasked with negotiating on behalf of individuals. By comparing agentic interactions to a human-to-human benchmark, the study reveals that instructional prompts act as carriers for the principal's personality, biases, and demographic traits. Remarkably, delegating decisions to machines leads to a greater

A Unifying View of Attention Sinks: Two Algorithms, Two Solutions Jun 16, 2026 00:22:35 This research investigates the nature of attention sinks, which are specific tokens in Transformer models that attract disproportionate attention. The authors reveal that these identical visual patterns actually facilitate two distinct computational algorithms: Adaptive NOP and Broadcast. In the Adaptive NOP mechanism, the model uses a "null" token with near-zero value to suppress update

From AGI to ASI Jun 14, 2026 00:23:37 This report from Google DeepMind explores the hypothetical transition from Artificial General Intelligence (AGI), which matches human capability, to Artificial Superintelligence (ASI), which far exceeds it. The authors outline four primary technological pathways to achieve this: quantitative scaling, algorithmic paradigm shifts, recursive self-improvement, and multi-agent coordination. While curre

Correct Looks Better: Pairwise Comparisons Reveal Accuracy Rankings Jun 13, 2026 00:19:52 This research explores whether pairwise comparisons used to rank generative models actually reflect ground-truth accuracy. By converting multiple benchmarks into free-form formats, the authors found that Elo-style rankings achieve a remarkably high correlation with objective correctness. Surprisingly, this alignment remains strong even when the judge model is weaker than the candidates it evaluate

Critical Batch Size for LLM Policy Optimization Jun 11, 2026 00:18:41 This paper investigates the critical batch size (CBS) for Large Language Model (LLM) policy optimization, specifically focusing on the GRPO algorithm. The researchers break down gradient noise into inter-prompt and intra-prompt components to determine the point where increasing data parallelism yields diminishing returns. Their findings reveal that on-policy training is primarily limited by noise

Self-supervised User Profile Generation for Personalization Jun 9, 2026 00:22:05 This paper describes a self-supervised framework called BUMP, which is designed to improve how large language models deliver personalized content. Traditionally, creating user profiles for search and recommendation tasks requires expensive, human-labeled data to train the system. To solve this, researchers developed a method that uses a bidirectional ranking objective to learn directly from raw in

From Augmentation to Reconstruction: Guiding the AI Disruption to the Good Place Jun 7, 2026 00:22:10 This paper explores the evolution of artificial intelligence through a three-stage framework of augmentation, automation, and reconstruction. The authors argue that while AI currently improves individual tasks, the most profound economic disruption will only occur when workflows and markets are entirely redesigned around machine capabilities. True transformation is currently stalled by legacy huma

Self-Distilled Agentic Reinforcement Learning Jun 7, 2026 00:22:14 The research paper introduces SDAR (Self-Distilled Agentic Reinforcement Learning), a new framework designed to improve the training of large language model agents in complex, multi-turn environments. While standard reinforcement learning excels at high-level task goals, it often lacks the precise, token-level guidance needed for long interactions. To solve this, the authors identify critical flaw

Subliminal Learning Is Steering Vector Distillation Jun 5, 2026 00:23:29 This research explores subliminal learning, a phenomenon where a student language model inherits behavioral traits from a teacher model even when trained on semantically unrelated data. The authors demonstrate that this process is driven by steering vector distillation, where the teacher’s system prompt acts as a linear direction in activation space that the student internalizes during fine-tuning

Subsidizing Sequential Search Jun 5, 2026 00:20:20 This paper explores a market model where competing firms use subsidies to reduce the cost of product inspection for consumers. Through a subsidy-sorting principle, the authors demonstrate that higher-quality firms naturally offer larger subsidies to signal their value and secure priority in the search order. This behavior results in a unique equilibrium where low-quality firms are ignored, interme

Meta-Harness: End-to-End Optimization of Model Harnesses Jun 2, 2026 00:17:39 This paper introduces Meta-Harness, an innovative system designed to automate harness engineering for large language models. Unlike traditional methods that rely on manual coding or compressed feedback, this system uses an agentic proposer to search through and optimize the code that governs how models store, retrieve, and process information. By utilizing a filesystem to access full execution tra

Self-Improving Language Models with Bidirectional Evolutionary Search Jun 1, 2026 00:20:59 Researchers have developed Bidirectional Evolutionary Search (BES) to overcome the limitations of standard language model sampling, which often struggles with sparse feedback and predictable outputs. While traditional methods like tree search are confined to a narrow "entropy shell" of high-probability responses, BES escapes this range by using evolutionary operators such as crossover an

Generative Modeling via Drifting May 31, 2026 00:21:53 This paper discusses Drifting Models, a novel generative modeling paradigm that enables high-quality, one-step image generation without the iterative inference required by diffusion or flow-matching models. Instead of decomposing transformations at the sampling stage, this method evolves a pushforward distribution during the training process by utilizing a neural network optimizer. The core mechan

Instance-Optimal Estimation with Multiple LLM Judges on a Budget May 31, 2026 00:21:23 This paper addresses the cost-efficient evaluation of large language models (LLMs) by utilizing multiple AI "judges" with different price points and reliability levels. The researchers formalize this challenge as budgeted heteroskedastic multi-judge estimation, seeking an optimal way to distribute a limited budget across various judges and tasks to achieve the most accurate quality score

Robust AI Personalization Will Require a Human Context Protocol May 29, 2026 00:22:59 This paper proposes the Human Context Protocol (HCP), a technical framework designed to give individuals direct control over how their personal preferences shape AI interactions. Currently, AI personalization relies on fragmented data silos and behavioral inferences that often fail to reflect a user’s true intent or values. By establishing a user-owned preference layer, the protocol allows people

Equilibrium Reasoners: Learning Attractors Enables Scalable Reasoning May 27, 2026 00:17:42 This paper introduces Equilibrium Reasoners (EqR), a novel framework that conceptualizes iterative AI reasoning as a dynamical system converging toward stable latent attractors. By treating the reasoning process as a series of repeated updates to an internal state, the researchers demonstrate that models can scale performance at test-time by simply increasing the number of iterations (depth) or us

Position: The Pre/Post-Training Boundary Should Govern IP in Industry–Academia ML Collaborations May 25, 2026 00:12:57 This paper proposes a new contractual framework called PBOS to resolve persistent intellectual property conflicts in industry-academia machine learning collaborations. By involving scientists in legal negotiations, the authors suggest a clear division based on the pre/post-training boundary of a model. Under this model, pre-training artifacts such as code and architectures are treated as open scie

Episodes

Recommended