Formal & Physical Sciences AI — Models & ResearchENSamyak Jain: Designing a Neural CompressorSamyak JainAugust 6, 2026 26 min★ ★ ★ ★ ☆ 4/5Neural CompressionTransformersArithmetic Coding
Formal & Physical Sciences Language & LinguisticsENMiloš Stanojević: Syntactic Belief Update as the Driver of Garden Path Processing DifficultyMiloš StanojevićAugust 6, 2026 53 min★ ★ ★ ★ ☆ 4/5Syntactic ProcessingBelief UpdateGarden Path
Formal & Physical Sciences AI — Models & ResearchENYash Sarrof: Length Generalization of Transformers with a Growing Test time AlphabetYash SarrofJuly 23, 2026 44 min★ ★ ★ ★ ☆ 4/5Length GeneralizationTransformersRASP
Formal & Physical Sciences Computing & CybersecurityENYang Wang: On regular copying languagesYang WangJuly 23, 2026 43 min★ ★ ★ ★ ☆ 4/5Formal LanguagesReduplicationFinite Automata
Formal & Physical Sciences AI — Models & ResearchENSelim Jerad: Context-Free Recognition with TransformersSelim JeradJuly 23, 2026 38 min★ ★ ★ ★ ☆ 4/5TransformersContext-Free GrammarsFormal Languages
Formal & Physical Sciences AI — Models & ResearchENPeter Shaw: Asymptotically Optimal Description Length Objectives for TransformersPeter ShawJuly 23, 2026 42 min★ ★ ★ ★ ☆ 4/5TransformersMinimum Description LengthAlgorithmic Information Theory
Formal & Physical Sciences AI — Models & ResearchENLimits of Deep Learning: Sequence Modeling through the Lens of Complexity TheoryNikola ZubicJuly 23, 2026 49 min★ ★ ★ ★ ☆ 4/5Complexity TheorySequence ModelsState Space Models
Formal & Physical Sciences AI — Models & ResearchENIdan Attias: On the Hardness of Learning Regular ExpressionsIdan AttiasJuly 23, 2026 56 min★ ★ ★ ★ ☆ 4/5Regular ExpressionsComputational Learning TheoryPAC Learning
Formal & Physical Sciences AI — Models & ResearchENChenxiao Yang: On Powerful Ways to Generate: Autoregression, Diffusion, and BeyondChenxiao YangJuly 23, 2026 58 min★ ★ ★ ★ ☆ 4/5Diffusion ModelsAutoregressive ModelsComputational Complexity
Formal & Physical Sciences Computing & CybersecurityENClemente Pasti: Prefix Parsing is Just ParsingClemente PastiJuly 2, 2026 75 min★ ★ ★ ★ ☆ 4/5Prefix ParsingContext-Free GrammarsWeighted Grammars
Formal & Physical Sciences AI — Models & ResearchENAleksandra Bakalova: Discovering Interpretable Algorithms by Decompiling Transformers to RASPAleksandra Bakalova, Shinting HuangJuly 2, 2026 39 min★ ★ ★ ★ ☆ 4/5TransformersMechanistic InterpretabilityRASP
Formal & Physical Sciences AI — Models & ResearchENHanlin Zhu: Reasoning by Superposition: A Theoretical Perspective on Chain of Continuous ThoughtHanlin ZhuMay 13, 2026 49 min★ ★ ★ ★ ☆ 4/5Chain-of-ThoughtContinuous ThoughtTransformer
Formal & Physical Sciences AI — Models & ResearchENAleksandar Terzić: Structured Sparse Transition Matrices to Enable State Tracking in SSMsAleksandar TerzićMay 13, 2026 50 min★ ★ ★ ★ ☆ 4/5State Space ModelsExpressivityFormal Languages
Formal & Physical Sciences AI — Models & ResearchENFrancesco Cagnetta: Deep Networks Learn to Parse Context-Free Languages from Local StatisticsFrancesco CagnettaApril 1, 2026 48 min★ ★ ★ ★ ☆ 4/5Deep LearningContext-Free GrammarsSample Complexity
Formal & Physical Sciences Language & LinguisticsENAdam Jardine: A weak but flexible logic for modeling phonological processesAdam JardineApril 1, 2026 47 min★ ★ ★ ★ ☆ 4/5PhonologyFormal Language TheorySubsequential Functions
Formal & Physical Sciences AI — Models & ResearchENAwni Altabaa: Unlocking Out-of-Distribution Generalization in Transformers via Latent ReasoningAwni AltabaaFebruary 28, 2026 49 min★ ★ ★ ★ ☆ 4/5TransformersOut-of-Distribution GeneralizationLatent Reasoning
Formal & Physical Sciences Language & LinguisticsENModelling a Subregular Bias in Phonological Learning with RNNsBrandon PrickettJanuary 31, 2026 43 min★ ★ ★ ★ ☆ 4/5PhonologyRNNSubregular
Formal & Physical Sciences AI — Models & ResearchENCai Zhou: Coevolutionary Continuous Discrete DLMs, Semantic Scale Prediction via Hierarchical DLMsCai ZhouJanuary 31, 2026 46 min★ ★ ★ ★ ☆ 4/5Diffusion Language ModelsHierarchical DiffusionContinuous-Discrete Diffusion
Formal & Physical Sciences AI — Models & ResearchENParsa Mirtaheri: Let Me Think! A Long Chain-of-Thought Can Be Worth Exponentially Many Short OnesParsa MirtaheriNovember 30, 2025 39 min★ ★ ★ ★ ☆ 4/5Chain-of-ThoughtInference-Time ScalingTransformers
Formal & Physical Sciences AI — Models & ResearchENLanguage Identification, Generation, and Hallucination Detection in the LimitGrigoris VelegkasNovember 17, 2025 49 min★ ★ ★ ★ ☆ 4/5Language IdentificationLanguage GenerationHallucination Detection
Formal & Physical Sciences AI — Models & ResearchENKulin Shah: Understanding Token Ordering in Masked DiffusionsKulin ShahNovember 17, 2025 48 min★ ★ ★ ★ ☆ 4/5Diffusion ModelsAutoregressive ModelsToken Ordering
Formal & Physical Sciences AI — Models & ResearchENDana Angluin: Trying to Understand TransformersDana AngluinNovember 17, 2025 58 min★ ★ ★ ★ ☆ 4/5TransformersExpressivityComplexity Theory
Formal & Physical Sciences AI — Models & ResearchENAwni Altabaa: CoT Information: Improved Sample Complexity under Chain-of-Thought SupervisionAwni AltabaaNovember 17, 2025 53 min★ ★ ★ ★ ☆ 4/5Chain-of-ThoughtSample ComplexityStatistical Learning Theory
Formal & Physical Sciences AI — Models & ResearchENAditya Varre: Learning In-context n-grams with Transformers: Sub-n-grams Are Near-stationary PointsAditya VarreOctober 26, 2025 48 min★ ★ ★ ★ ☆ 4/5In-Context LearningTransformersN-Gram Models
Formal & Physical Sciences AI — Models & ResearchENYing Fan: Looped Transformers for Length GeneralizationYing FanOctober 3, 2025 33 min★ ★ ★ ★ ☆ 4/5Length GeneralizationLooped TransformersTransformers
Formal & Physical Sciences AI — Models & ResearchENEshaan Nichani: Learning Compositional Functions with Transformers from Easy-to-Hard DataEshaan NichaniOctober 3, 2025 46 min★ ★ ★ ★ ☆ 4/5TransformersCompositional FunctionsCurriculum Learning
Formal & Physical Sciences AI — Models & ResearchENEkdeep Lubana: In-context Learning of Formal LanguagesEkdeep LubanaOctober 3, 2025 48 min★ ★ ★ ★ ☆ 4/5In-Context LearningFormal LanguagesTransformers
Formal & Physical Sciences AI — Models & ResearchENAndrew Gordon Wilson: Deep Learning is Not So Mysterious or DifferentAndrew Gordon WilsonOctober 3, 2025 56 min★ ★ ★ ★ ☆ 4/5Deep LearningGeneralizationPAC-Bayes
Formal & Physical Sciences NeuroscienceENMichael Goodale: Meta-Learning Neural Mechanisms rather than Bayesian PriorsMichael GoodaleOctober 1, 2025 36 min★ ★ ★ ★ ☆ 4/5Meta-LearningBayesian PriorsNeural Mechanisms
Formal & Physical Sciences AI — Models & ResearchENAshok Vardhan: Attention with Markov: A Curious Case of Single-layer TransformersAshok VardhanOctober 1, 2025 37 min★ ★ ★ ★ ☆ 4/5TransformersMarkov ChainsAttention
Formal & Physical Sciences AI — Models & ResearchENThomas Chen: Non-Asymptotic Length GeneralizationThomas ChenAugust 26, 2025 50 min★ ★ ★ ★ ☆ 4/5Length GeneralizationTransformersRASP
Formal & Physical Sciences AI — Models & ResearchENNirmit Joshi: A Theory of Learning with Autoregressive Chain of ThoughtNirmit JoshiAugust 26, 2025 51 min★ ★ ★ ★ ☆ 4/5Chain of ThoughtLearning TheorySample Complexity
Formal & Physical Sciences AI — Models & ResearchENJuno Kim: Transformers Provably Solve Parity Efficiently with Chain of ThoughtJuno KimAugust 26, 2025 47 min★ ★ ★ ★ ☆ 4/5TransformersChain of ThoughtParity
Formal & Physical Sciences AI — Models & ResearchENDeqing Fu: Algorithmic Perspectives on Understanding TransformersDeqing FuAugust 26, 2025 45 min★ ★ ★ ★ ☆ 4/5TransformersIn-Context LearningNewton's Method
Formal & Physical Sciences AI — Models & ResearchENTraining Neural Networks as Recognizers of Formal LanguagesAlexandra ButoiAugust 18, 2025 29 min★ ★ ★ ★ ☆ 4/5Formal LanguagesNeural NetworksRNN
Formal & Physical Sciences AI — Models & ResearchENRuiquan Huang: A Theoretical Study on Training Dynamics and Implicit BiasRuiquan HuangAugust 18, 2025 51 min★ ★ ★ ★ ☆ 4/5TransformersTraining DynamicsImplicit Bias
Formal & Physical Sciences AI — Models & ResearchENNoPE: The Counting Power of Transformers with No Positional EncodingsChris KöcherAugust 18, 2025 58 min★ ★ ★ ★ ☆ 4/5TransformersFormal LanguagesCounting
Formal & Physical Sciences AI — Models & ResearchENJiaoda Li: Characterizing the Expressivity of Transformer Language ModelsJiaoda LiAugust 18, 2025 40 min★ ★ ★ ★ ☆ 4/5TransformersExpressivityFormal Languages