What the smart people are recommending. 8874 books, tools, and products endorsed by the thinkers absorb.md tracks. Ranked by how many times each has been recommended across compiled podcasts, papers, posts, and tweets.
“We release the trained weights in ONNX format under a non-commercial licence, with local desktop and real-time camera support, aimed explicitly at ecologists with no machine-learning experience.”
“there's an Incredible video that carpath just dropped that Andre just dropped where he does like this deep dive into llms and he explains chat GPT from the ground up it's on YouTube it's three hours i…”
“Through experiments on mathematical reasoning, code generation, and summarization, we show that agreement in representation space provides a robust and scalable signal of self-consistency for open-end…”
“Participating systems primarily fine-tuned Arabic and multilingual transformer-based models, including MARBERT, AraBERT, and DeBERTa-v3 variants, with several teams employing cross-validation, ensembl…”
“Participating systems primarily fine-tuned Arabic and multilingual transformer-based models, including MARBERT, AraBERT, and DeBERTa-v3 variants, with several teams employing cross-validation, ensembl…”
“Participating systems primarily fine-tuned Arabic and multilingual transformer-based models, including MARBERT, AraBERT, and DeBERTa-v3 variants, with several teams employing cross-validation, ensembl…”
“the dolly Lama has these great on his YouTube channel he's got these great like 2hour discussions you get about 20 30 minutes into that you will fall asleep”
“what we're doing right now ladies and gentlemen is uh sexy voice sexy mode grock Ai”
“For timing detection, we propose MEGUMI, a Multilingual Emotion-aware Gated Unit for Mutual Integration, that fuses frozen XLM-RoBERTa semantics with language-specific emotion encoders via cross-modal…”
“Claude is the best. I mean 3.7 is is really the best at doing agent stuff.”
“Finally, our EmoValidBench benchmarks of GPT-4.1 Nano and Llama-3.1 8B indicate that current LLMs generate contextually similar and diverse validating responses, but emotional understanding remains a …”
“Finally, our EmoValidBench benchmarks of GPT-4.1 Nano and Llama-3.1 8B indicate that current LLMs generate contextually similar and diverse validating responses, but emotional understanding remains a …”
“To address this dilemma, we propose Anchored Privacy Drifting (APD), a training-free method that drifts privacy-sensitive elements toward semantically equivalent alternatives while anchoring contextua…”
“To systematically evaluate this dual objective of privacy protection and contextual preservation, we introduce AdaptShield, a comprehensive benchmark covering 22 privacy categories, which combines con…”
“We present 3DINO, a cutting-edge SSL method adapted to 3D datasets, and use it to pretrain 3DINO-ViT: a general-purpose medical imaging model, on an exceptionally large, multimodal, and multi-organ da…”
“we've been working on this model for a while...we train Claud 3.7 you know more to focus on these real world tasks”
“We present 3DINO, a cutting-edge SSL method adapted to 3D datasets, and use it to pretrain 3DINO-ViT: a general-purpose medical imaging model, on an exceptionally large, multimodal, and multi-organ da…”
“Here we present CaliPPer (Calibration and Prediction of Performance), a post-hoc framework pairing a multi-chain Sample-to-Domain Distance (S2DD) with distance-aware Bayesian recalibration, operating …”
“Density-ratio methods (PAPE, M-CBPE) provide label-free performance estimation for binary classification, but their assumptions and aggregate-only outputs limit binding prediction on neoepitopes, anti…”
“Density-ratio methods (PAPE, M-CBPE) provide label-free performance estimation for binary classification, but their assumptions and aggregate-only outputs limit binding prediction on neoepitopes, anti…”
“A simplified rigid-beam model qualitatively captures the crossover behavior.”
“Beyond accuracy, human-annotated analyses show that DiScO substantially improves the model's ability to recover from erroneous initial attempts.”
“Experimental results demonstrate that our model, named Lius, yields notable improvements over standard instruction-tuned models by outperforming 4-6 points, and surpassing both Neural Machine Translat…”
“I'm on X I'm always on X I'm following 26,000 people in the AI Community I've built a bunch of list of 26,000 people and 6,000 companies on X so you can watch my list and join in the fun”
“I use XPro so I can see uh I don't know 10 different lists on my screen at one time and so I can watch literally the world is just going by”
“we present AutoPilot, a reinforcement learning-based framework that continuously monitors runtime conditions and dynamically adjusts protocol parameters online to optimize consensus performance.”
“We implement AutoPilot on top of Autobahn, a state-of-the-art, highspeed, robust BFT protocol, and evaluate it across diverse dynamic environments.”
“As a result, our SCPI classifier can effectively identify information related to SCPI.”
“We released this thing called the Anthropic Economic Index that, in a privacy-preserving way, looks through, you know, and summarizes our usage to understand, you know, in what fields are people using…”
“The results demonstrate that IARE achieves state-of-the-art performance while also generating accurate rationales.”
“MAVIS first bridges the granularity mismatch by parsing raw videos into a \textbf{Structured Semantic Library}, enabling explicit attribute-level indexing.”
“Crucially, MAVIS employs a \textbf{Logic-aware Debate} mechanism with a strict veto protocol, where agents collaboratively prune logical mismatches to identify a compact set of ``controversial'' candi…”
“A framework, MAFS, is introduced that provides an approximate description of the hierarchies of quark and lepton masses and mixing angles in terms of a set of small parameters, $ε_a$, one for each fer…”
“Introducing the LangGraph Functional API”
“Compared to a simpler model that only used the question stem and the key, our best distractor-aware architecture achieved higher predictive performance”
“An order-invariant variant achieved nearly the same accuracy with approximately half as many parameters, offering a favorable accuracy-efficiency trade-off”
“we propose a novel end-to-end semantic change detection network named SemDINO”
“ShallowBench therefore provides a rigorous benchmark for generative biology models and highlights the necessity of new architectural innovations or loss functions capable of navigating these challengi…”
“We evaluate the proposed flow using a Tensor Slice-based FPGA architecture with AMD Vitis HLS and the Verilog-to-Routing (VTR) toolchain.”
“We evaluate the proposed flow using a Tensor Slice-based FPGA architecture with AMD Vitis HLS and the Verilog-to-Routing (VTR) toolchain.”
“We propose a framework, called piecewise-linear Karush--Kuhn--Tucker hard-constrained PINNs (PL-KKT-hPINNs), that strictly enforces nonlinear equality constraints through piecewise-linear projection.”
“This extends the KKT-hPINN framewor, which exactly enforces linear equalities through the Karush--Kuhn--Tucker (KKT) conditions associated with orthogonally projecting neural network outputs onto the …”
“We observe that the strongest agents, Claude Opus 4.6 and GPT-5.4 xhigh, often avoid writing the target language directly.”
“We observe that the strongest agents, Claude Opus 4.6 and GPT-5.4 xhigh, often avoid writing the target language directly.”
“In contrast, Opus-derived Python helper code for building generators, with no solved benchmark programs or hidden-test answers, sharply improves Sonnet 4.6 and GPT-5.4 mini on the same problems, while…”
“In contrast, Opus-derived Python helper code for building generators, with no solved benchmark programs or hidden-test answers, sharply improves Sonnet 4.6 and GPT-5.4 mini on the same problems, while…”
“This article uses Moonshine's exploration of the Jacobian conjecture as an example.”
“By invoking GPT-5.5-pro and DeepSeek-V4-pro separately, Moonshine obtained independent complete proofs for the case \(N=n+1\).”
“By invoking GPT-5.5-pro and DeepSeek-V4-pro separately, Moonshine obtained independent complete proofs for the case \(N=n+1\).”
“This work presents Earth-OneVision, a 2B RS-MLLM that unifies six sensor modalities (i.e., optical, SAR, infrared, multispectral, temporal, and video) and cross-sensor fusion across 9 task categories …”
“To support joint training, MMRS-OneVision is constructed with ~34M QA pairs spanning all six sensor modalities and cross-sensor fusion across 9 task categories, substantially exceeding existing RS mul…”
“We propose ADAS, a training-free reranking rule for parallel masked diffusion decoding.”
“there's so many open source packages go start programming quantum computer go start learning and ultimately just do research and you'll quickly get into the field”
“We present LogNEO, a log anomaly detector built on EleutherAI's GPT-Neo (1.3B parameters) and fine-tuned with a novel partial-credit, exponentially decaying position-aware reward scheme combined with …”
“built on EleutherAI's GPT-Neo (1.3B parameters)”
“A production microservice deployment over Apache Kafka, Redis, and TensorRT-accelerated inference demonstrates 45 ms end-to-end latency at 15,000 events per second.”
“A production microservice deployment over Apache Kafka, Redis, and TensorRT-accelerated inference demonstrates 45 ms end-to-end latency at 15,000 events per second.”
“A production microservice deployment over Apache Kafka, Redis, and TensorRT-accelerated inference demonstrates 45 ms end-to-end latency at 15,000 events per second.”
“As the Model Context Protocol (MCP) standardizes tool-calling for autonomous agents, it introduces a critical, unexamined attack surface: the error-handling loop.”
“Our evaluation across four frontier models, Gemini 3.1 Pro, GPT-5.5, GLM-5.1, and Qwen3-Coder, demonstrates that error-path injection triples the success rate of standard indirect prompt injection (IP…”
“Our evaluation across four frontier models, Gemini 3.1 Pro, GPT-5.5, GLM-5.1, and Qwen3-Coder, demonstrates that error-path injection triples the success rate of standard indirect prompt injection (IP…”
“When properly implemented and optimized, PO4AO constitutes a robust and high-performance turnkey controller for single-conjugate adaptive optics systems,...”
“The nnU-Net has demonstrated continuous success in medical segmentation tasks”
“in this course we'll be using a package called langme um which contains a bunch of utilities... it contains utilities for when it's updating to semantic memory”
“we propose CITRAS-FM, a tiny 7M-parameter TSFM that supports univariate, multivariate, and covariate-informed zero-shot forecasting with real-time CPU inference.”
“we propose CovSynth, which synthesizes realistic covariates from decomposed components of target series.”
“one of the things that we actually use is we use this thing called trust call. Um this thing called trust call basically applies large JSON schema updates to things.”
“We propose a deep reinforcement learning framework for multi-objective policy optimization at this scale.”
“Introducing MIST: a benchmark of synthetically generated multi-turn conversations where users express plausible misconceptions in scientific, medical, and moral reasoning domains.”
“We propose DAC (Divide and Cooperate), a role-decomposed multi-agent training framework that divides agentic search into two cooperative subtasks, each handled by a dedicated agent trained with role-s…”
“use replet agent even I would say openi deep research will give you an idea of like you know where these tools are coming going”
“LLM-powered chatbots are increasingly embedded in everyday workflows, raising sustainability concerns due to their energy use.”
“I started like playing around with Gemini from Google I would say Gemini Flash is incredibly impressive Really large context window”
“We then evaluated the prototype in a five-day field study with 11 participants.”
“We present a lightweight, generative approach to provenance tracking based on observational semantics. Instead of propagating identifiers through compiler passes, we observe graph transformations and …”
“Foreground Conditioned Outpainting (FCO) meets this demand, allowing users to create desired backgrounds for foreground instances at a low cost by adjusting the text prompt.”
“We then introduce Gravity-Weighted DPO (GW-DPO), a preference-optimization objective whose per-sample offset scales with the structural distance between conflicting levels under a linear or bilateral …”
“Combined with hierarchy-specific delimiter tokens (Chen et al., 2025) and Instructional Segment Embeddings (ISE; Wu et al., 2025), GW-DPO with the bilateral schedule Pareto-improves over standard DPO …”
“There's a bunch of websites there's composio um a lot of startups I think another one called uh tool MCP tools or something like that Smither AI and other ones provide you model context protocols”
“sonnet 3.7 claw sonnet 3.7 you know everyone who's using cursor whatever thinks it's bad... what I saw in that is an AI that can step back and think... I think what anthropic did is they gave it agenc…”
“We revisit graph-based clustering as a bottom-up alternative, where segment embeddings are connected by pairwise similarity and partitioned using the Leiden algorithm.”
“Another bottom-up approach, agglomerative clustering with average linkage, also performs well, although it is computationally less efficient and allows for less control over the resulting distribution…”
“We describe ConvMemory v2, an opt-in token-evidence reranker that sits after the lightweight ConvMemory v1 reranker and reorders only v1's protected top-10 candidate set.”
“We describe ConvMemory v2, an opt-in token-evidence reranker that sits after the lightweight ConvMemory v1 reranker and reorders only v1's protected top-10 candidate set.”
“v2 is a fine-tuned ms-marco-MiniLM-L-6-v2 cross-encoder (22,713,601 parameters, measured from the released checkpoint) applied to the ten (query, memory) pairs that v1 has already selected; it does no…”
“On the LoCoMo conversational memory benchmark (5 seeds, n = 4955 test rows), v2 raises FULL MRR from v1's 0.5824 to 0.6560 (paired bootstrap +0.0734, 95% CI [+0.0645, +0.0827]) and H@1 from 0.4440 to …”
“To address these limitations, we propose DocTrace, a multi-agent RAG framework for long-document QA that supports query-triggered knowledge organization, document-structure-aware and experience-guided…”
“Based on these insights, we propose the Knowledge-Augmented Tool Execution (KATE), a knowledge-augmented tool execution framework that integrates experiential knowledge with reasoning-width-expanded i…”
“we propose AutoPDE, a code agent that maintains the solver strategy as an explicitly represented object throughout the solving process”
“Specifically, experiments with popular methods like ROME and FT reveal a substantial performance gap, up to 24%, between evaluations on directly edited knowledge and on entailed knowledge.”
“Specifically, experiments with popular methods like ROME and FT reveal a substantial performance gap, up to 24%, between evaluations on directly edited knowledge and on entailed knowledge.”
“If who have you have heard about slide the presentation framework um no one you should check it out it's cool.”
“we propose Causal Ensemble Agent (CEA), a novel framework that aggregates structural insights from statistical discovery experts across different graph levels via linear opinion pooling, and uses an L…”
“We propose FlowTracer, an RL framework that traces answer-targeted reasoning flow on an attention-induced directed acyclic graph”
“In this work, we propose the Unified Dexterous Hand Model (UDHM), which maps human and robot hand states into a shared 22-DoF semantic interface.”
“In this work, we propose the Unified Dexterous Hand Model (UDHM), which maps human and robot hand states into a shared 22-DoF semantic interface.”
“Ultimately, ParaEval provides a robust and efficient way to evaluate true underlying capability rather than surface-form familiarity.”
“We propose Spense, a practical hybrid sparse-dense format that splits each weight matrix into a 2:4 sparse region and a dense region. This design relaxes the effective sparsity constraint while remain…”
“We propose Spense, a practical hybrid sparse-dense format that splits each weight matrix into a 2:4 sparse region and a dense region.”
“We propose Prefilling-dLLM, a training-free prefill-decode disaggregation framework for dLLMs that partitions the prefix into N chunks, caches their KV representations once, and selects the top-K most…”