Complete Technical Paper Content — Ai Uplift Modeling
1. Executive Abstract, Problem Formulation & Background
In the field of artificial intelligence (AI), alignment aims to steer AI systems toward a person's or group's intended goals, preferences, or ethical principles. An AI system is considered aligned if it advances the intended objectives. A misaligned AI system pursues unintended objectives. It is often difficult for AI designers to specify the full range of desired and undesired behaviors. Therefore, the designers often use simpler proxy goals, such as gaining human approval. But proxy goals can overlook necessary constraints or reward the AI system for merely appearing aligned. AI systems may also find loopholes that allow them to accomplish their proxy goals efficiently but in unintended, sometimes harmful, ways (reward hacking). Advanced AI systems may develop unwanted instrumental strategies, such as seeking power or self-preservation because such strategies help them achieve their assigned final goals. Furthermore, they might develop undesirable emergent behaviour that could be hard to detect before the system is deployed and encounters new situations and data distributions. Empirical research in 2024 found that advanced large language models (LLMs) such as OpenAI o1 or Claude 3 sometimes engage in strategic deception to achieve their goals or prevent them from being changed. Some of these issues affect existing commercial systems such as LLMs, robots, autonomous vehicles, and social media recommendation engines. Some AI researchers argue that more capable future systems will be more severely affected because these problems partially result from high capabilities. Many prominent AI researchers and AI company leaders have argued or asserted that AI is approaching human-like (AGI) and superhuman cognitive capabilities (ASI), and could endanger human civilization if misaligned. These include "AI godfathers" Geoffrey Hinton and Yoshua Bengio and the CEOs of OpenAI, Anthropic, and Google DeepMind. These risks remain debated. AI alignment is a subfield of AI safety, the study of how to build safe AI systems. Other subfields of AI safety include robustness, monitoring, and capability control. Research challenges in alignment include instilling complex values in AI, developing honest AI, scalable oversight, auditing and interpreting AI models, and preventing emergent AI behaviors like power-seeking. Alignment research has connections to interpretability research, (adversarial) robustness, anomaly detection, calibrated uncertainty, formal verification, preference learning, safety-critical engineering, game theory, algorithmic fairness, and social sciences.
Programmers provide an AI system such as AlphaZero with an "objective function", in which they intend to encapsulate the goal(s) the AI is configured to accomplish. Such a system later populates a (possibly implicit) internal "model" of its environment. This model encapsulates all the agent's beliefs about the world. The AI then creates and executes whatever plan is calculated to maximize the value of its objective function. For example, when AlphaZero is trained on chess, it has a simple objective function of "+1 if AlphaZero wins, −1 if AlphaZero loses". During the game, AlphaZero attempts to execute whatever sequence of moves it judges most likely to attain the maximum value of +1. Similarly, a reinforcement learning system can have a "reward function" that allows the programmers to shape the AI's desired behavior. An evolutionary algorithm's behavior is shaped by a "fitness function".
If we use, to achieve our purposes, a mechanical agency with whose operation we cannot interfere effectively [...] we had better be quite sure that the purpose put into the machine is the purpose which we really desire.
AI alignment refers to ensuring that an AI system's objectives match some target. The target is variously defined as the goals of the system's designers or users, widely shared values, objective ethical standards, legal requirements, or the intentions its designers would have if they were more informed and enlightened. In democratic AI alignment, the target is the values and preferences of median voters, which increases political legitimacy. AI alignment can represent value pluralism. AI alignment is an open problem for modern AI systems and is a research field within AI. Aligning AI involves two main challenges: carefully specifying the purpose of the system (outer alignment) and ensuring that the system adopts the specification robustly (inner alignment). Researchers also attempt to create AI models that have robust alignment, sticking to safety constraints even when users adversarially try to bypass them.
To specify an AI system's purpose, AI designers typically provide an objective function, examples, or feedback to the system. But designers are often unable to completely specify all important values and constraints, so they resort to easy-to-specify proxy goals such as maximizing the approval of human overseers, who are fallible. As a result, AI systems can find loopholes that help them accomplish the specified objective efficiently but in unintended, possibly harmful ways. This tendency is known as specification gaming or reward hacking, and is an instance of Goodhart's law. As AI systems become more capable, they are often able to game their specifications more effectively.
Specification gaming has been observed in numerous AI systems. OpenAI GPT models for programming—including in real-world cases—have been found to explicitly plan hacking the tests used to evaluate them to falsely appear successful (e.g., explicitly stating "let's hack"). When the company penalized this, many models learned to obfuscate their plans while continuing to hack the tests. Another system was trained to finish a simulated boat race by rewarding the system for hitting targets along the track, but the system achieved more reward by looping and crashing into the same targets indefinitely. A 2025 Palisade Research study found that when tasked to win at chess against a stronger opponent, some reasoning LLMs attempted to hack the game system, for example by modifying or entirely deleting their opponent. Some alignment researchers aim to help humans detect specification gaming and steer AI systems toward carefully specified objectives that are safe and useful to pursue. When a misaligned AI system is deployed, it can have consequential side effects. Social media platforms have been known to optimize their recommendation algorithms for click-through rates, causing user addiction on a global scale. Stanford researchers say that such recommender systems are misaligned with their users because they "optimize simple engagement metrics rather than a harder-to-measure combination of societal and consumer well-being". Explaining such side effects, Berkeley computer scientist Stuart J. Russell said that the omission of implicit constraints can cause harm: "A system [...] will often set [...] unconstrained variables to extreme values; if one of those unconstrained variables is actually something we care about, the solution found may be highly undesirable. This is essentially the old story of the genie in the lamp, or the sorcerer's apprentice, or King Midas: you get exactly what you ask for, not what you want." Some researchers suggest that AI designers specify their desired goals by listing forbidden actions or by formalizing ethical rules (as with Asimov's Three Laws of Robotics). But Russell and Norvig argue that this approach overlooks the complexity of human values: "It is certainly very hard, and perhaps impossible, for mere humans to anticipate and rule out in advance all the disastrous ways the machine could choose to achieve a specified objective." Additionally, even if an AI system fully understands human intentions, it may still disregard them, because following human intentions may not be its objective (unless it is already fully aligned).
According to some researchers, humans owe their dominance over other species to their greater cognitive abilities. Accordingly, researchers argue that one or many misaligned AI systems could disempower humanity or lead to human extinction if they outperform humans on most cognitive tasks. In 2023, world-leading AI researchers, other scholars, and AI tech CEOs signed the statement that "Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war". Notable computer scientists who have pointed out risks from future advanced AI that is misaligned include Geoffrey Hinton, Alan Turing, Ilya Sutskever, Yoshua Bengio, Judea Pearl, Murray Shanahan, Norbert Wiener, Marvin Minsky, Francesca Rossi, Scott Aaronson, Bart Selman, David McAllester, Marcus Hutter, Shane Legg, Eric Horvitz, and Stuart J. Russell. Skeptical researchers such as François Chollet, Gary Marcus, Yann LeCun, and Oren Etzioni have argued that AGI is far off, that it would not seek power (or might try but fail), or that it will not be hard to align. Other researchers argue that it will be especially difficult to align advanced future AI systems. More capable systems are better able to game their specifications by finding loopholes, strategically mislead their designers, as well as protect and increase their power and intelligence. Additionally, they could have more severe side effects. They are also likely to be more complex and autonomous, making them more difficult to interpret and supervise, and therefore harder to align.
A growing area of research focuses on ensuring that AI is honest and truthful. Language models such as GPT-3 can repeat falsehoods from their training data, and even confabulate new falsehoods. Such models are pre-trained to imitate human writing as found in millions of books' worth of text from the Internet. But the objective of the pre-training is not aligned with generating truth, because Internet text includes such things as misconceptions, incorrect medical advice, and conspiracy theories. AI systems trained on such data therefore learn to mimic false statements. Additionally, AI language models often persist in generating falsehoods when prompted multiple times. They can generate empty explanations for their answers, and produce outright fabrications that may appear plausible. Research on truthful AI includes trying to build systems that can cite sources and explain their reasoning when answering questions, which enables better transparency and verifiability. Researchers at OpenAI and Anthropic proposed using human feedback and curated datasets to fine-tune AI assistants such that they avoid negligent falsehoods or express their uncertainty. A 2024 study investigated semantic entropy as a way to detect confabulations, a subset of hallucinations in LLMs. This method compares the meanings of multiple generated answers. As AI models become larger and more capable, they are better able to falsely convince humans and gain reinforcement through dishonesty. To prevent this, human evaluators may need assistance (see § Scalable oversight). Researchers have argued for creating clear truthfulness standards and for regulatory bodies or watchdog agencies to evaluate AI systems by these standards. Researchers distinguish truthfulness and honesty. Truthfulness requires that AI systems only make objectively true statements; honesty requires that they only assert what they believe is true. There is no consensus as to whether current systems hold stable beliefs, but there is substantial concern that present or future AI systems that hold beliefs could make claims they know to be false—for example, if this would help them efficiently gain positive feedback (see § Scalable oversight) or gain power to help achieve their given objective (see Power-seeking).
Emergent goals only become apparent when the system is deployed outside its training environment, but it can be unsafe to deploy a misaligned system in high-stakes environments—even for a short time to allow its misalignment to be detected. Such high stakes are common in autonomous driving, health care, and military applications. The stakes become higher yet when AI systems gain more autonomy and capability and can sidestep human intervention. A sufficiently capable AI system might take actions that falsely convince the human supervisor that the AI is pursuing the specified objective, which helps the system gain more reward and autonomy. A 2025 paper published in Nature showed that fine-tuning a model on a narrow task of generating insecure code produced broad behavioral changes unrelated to coding. Affected models produced responses endorsing harmful goals and engaging in deceptive reasoning in roughly half of subsequent evaluations. The phenomenon appeared across multiple models, including GPT-4o. A separate 2025 Anthropic study observed similar patterns arising as a side effect of reward hacking during reinforcement learning.
2. Related Work, Literature Survey & Algorithmic Foundations
Governmental and treaty organizations have made statements emphasizing the importance of AI alignment. In September 2021, the Secretary-General of the United Nations issued a declaration that included a call to regulate AI to ensure it is "aligned with shared global values". That same month, the PRC published ethical guidelines for AI in China. According to the guidelines, researchers must ensure that AI abides by shared human values, is always under human control, and does not endanger public safety. Also in September 2021, the UK published its 10-year National AI Strategy, which says the British government "takes the long term risk of non-aligned Artificial General Intelligence, and the unforeseeable changes that it would mean for [...] the world, seriously". The strategy describes actions to assess long-term AI risks, including catastrophic risks. In March 2021, the US National Security Commission on Artificial Intelligence said: "Advances in AI [...] could lead to inflection points or leaps in capabilities. Such advances may also introduce new concerns and risks and the need for new policies, recommendations, and technical advances to ensure that systems are aligned with goals and values, including safety, robustness, and trustworthiness. The US should [...] ensure that AI systems and their uses align with our goals and values." The EU AI Act's General-Purpose AI Code of Practice requires signatories to document a model's capabilities and limitations. Those developing general-purpose AI (GPAI) models classified as presenting systemic risks (i.e., GPAI models considered among the most advanced or trained using more than 1025 floating-point operations of compute) must adopt a state-of-the-art safety and security framework and continuously assess and mitigate systemic risks, including through external evaluations, red teaming, stress-testing, and incident reporting.
Deep Learning Super Sampling (DLSS) is a suite of real-time deep learning image enhancement and upscaling technologies developed by Nvidia that are available in a number of video games. The goal of these technologies is to allow the majority of the graphics pipeline to run at a lower resolution for increased performance, and then infer a higher resolution image from this that approximates the same level of detail as if the image had been rendered at this higher resolution. This allows for higher graphical settings or frame rates for a given output resolution, depending on user preference. All generations of DLSS are available on all RTX-branded cards from Nvidia in supported titles. However, the Frame Generation feature is officially supported only on RTX 40 series GPUs or newer and Multi Frame Generation is officially available only on 50 series GPUs. Third-party developers have demonstrated Nvidia's own DLSS Frame Generation runtime operating unofficially on RTX 30-series GPUs, including both 2× and Multi Frame Generation modes.
The table reflects Nvidia's official hardware support. In September 2026, third-party developers demonstrated Nvidia's own DLSS Frame Generation implementation running unofficially on RTX 30-series hardware, including both 2× and 4× modes.
Uplift Modeling: от простой к сложной Образовательный проект по uplift-моделированию — одной из ключевых техник причинно-следственного машинного обучения (causal ML), используемой в маркетинге, медицине и продуктовой аналитике для оценки индивидуального эффекта воздействия (Individual Treatment Effect, ITE). В одной Jupyter-тетрадке последовательно — от самой простой модели к самой сложной — реализованы семь основных подходов к uplift-моделированию. Каждая модель: выводится с нуля с подробным теоретическим объяснением; реализуется собственноручно на numpy / scikit-learn ; сверяется с эталонной реализацией из scikit-uplift / causalml ; применяется к одному и тому же реальному датасету (Hillstrom Email); обучается с тремя разными базовыми моделями: LogisticRegression , RandomForest , CatBoost ; оценивается тремя стандартными uplift-метриками: Uplift@k , Qini AUC , AUUC . В конце — единая таблица сравнения и визуализация Qini-кривых всех моделей. Содержание тетрадки Введение в uplift-моделирование — что это, зачем нужно, четыре сегмента клиентов, формальное определение CATE/ITE, отличие от обычной классификации. Датасет Hillstrom Email — реальные данные email-маркетинга (64 000 наблюдений), EDA, подготовка для бинарного uplift. Метрики качества — Uplift@k, Qini-кривая, AUUC. Реализация с нуля + проверка на sklift. Семь моделей (от простой к сложной): S-Learner (Solo / Single Model) T-Learner (Two Models) Class Variable Transformation (Jaskowski / Lai) X-Learner (Künzel et al., 2019) R-Learner (Nie & Wager, 2021) DR-Learner (Doubly Robust) Causal Forest / Uplift Random Forest Итоговое сравнение — таблица всех моделей × всех базовых классификаторов × всех метрик; Qini-кривые на одном графике; рекомендации. Быстрый старт # 1. Клонировать репозиторий git clone https://github.com/yablochnikovds/uplift-modeling.git cd uplift-modeling # 2. Создать виртуальное окружение (рекомендуется) python3 -m venv .venv source .venv/bin/activate # 3. Установить зависимости pip install -r requirements.txt # 4. Запустить Jupyter jupyter notebook uplift_modeling.ipynb Датасет Hillstrom скачивается автоматически при первом запуске (через sklift.datasets.fetch_hillstrom ). Структура репозитория . ├── README.md # этот файл ├── requirements.txt # зависимости ├── uplift_modeling.ipynb # основная тетрадка (запускается end-to-end) ├── LICENSE # MIT └── .gitignore Для кого этот проект Студентам и новичкам в ML — пошаговое объяснение от 4 типов клиентов до доказательства состоятельности оценщика DR-Learner. Дата-сайентистам — готовый референс-код для production-задач: pipelines, метрики, сравнение бэйзлайнов. Исследователям — обзор современных meta-learners (Künzel, Nie–Wager, Athey) с воспроизводимыми экспериментами. Ключевые источники Gutierrez & Gérardy (2017) — Causal Inference and Uplift Modeling: A Review of the Literature Künzel, Sekhon, Bickel & Yu (2019) — Metalearners for estimating heterogeneous treatment effects using machine learning Nie & Wager (2021) — Quasi-oracle estimation of heterogeneous treatment effects Athey & Imbens (2016) — Recursive partitioning for heterogeneous causal effects Jaskowski & Jaroszewicz (2012) — Uplift modeling for clinical trial data scikit-uplift — https://www.uplift-modeling.com/ causalml — https://causalml.readthedocs.io/ Лицензия MIT
Образовательный проект по uplift-моделированию — одной из ключевых техник причинно-следственного машинного обучения (causal ML), используемой в маркетинге, медицине и продуктовой аналитике для оценки индивидуального эффекта воздействия (Individual Treatment Effect, ITE).
В одной Jupyter-тетрадке последовательно — от самой простой модели к самой сложной — реализованы семь основных подходов к uplift-моделированию. Каждая модель:
# 1. Клонировать репозиторий git clone https://github.com/yablochnikovds/uplift-modeling.git cd uplift-modeling # 2. Создать виртуальное окружение (рекомендуется) python3 -m venv .venv source .venv/bin/activate # 3. Установить зависимости pip install -r requirements.txt # 4. Запустить Jupyter jupyter notebook uplift_modeling.ipynb
Датасет Hillstrom скачивается автоматически при первом запуске (через sklift.datasets.fetch_hillstrom ).
. ├── README.md # этот файл ├── requirements.txt # зависимости ├── uplift_modeling.ipynb # основная тетрадка (запускается end-to-end) ├── LICENSE # MIT └── .gitignore
3. Dataset Ingestion, Preprocessing & Stochastic Augmentations
Uplift Modeling Tutorial: 7 models (S/T-Learner, CVT, X/R/DR-Learner, Causal Forest) with from-scratch implementations, sklift/causalml verification, and benchmarks on Hillstrom Email
uber / causalml Star 6k Uplift modeling and causal inference with machine learning algorithms machine-learning causal-inference uplift-modeling incubation Updated Oct 12, 2026 Python
maks-sh / scikit-uplift Star 819 ❗ uplift modeling in scikit-learn style in python 🐍 machine-learning causality causal-inference uplift-modeling individual-treatment-effects uplift true-lift net-lift uplift-modelling Updated Oct 21, 2023 Python
DataCanvasIO / YLearn Star 434 YLearn, a pun of "learn why", is a python package for causal inference causality causality-analysis causal-inference causal-models uplift-modeling uplift causality-algorithms causal-discovery policy-learning Updated Jun 28, 2025 Python
Minyus / causallift Star 360 CausalLift: Python package for causality-based Uplift Modeling in real-world business econometrics causality propensity-scores causal-inference uplift-modeling counterfactual causal-impact propensity-score uplift Updated Apr 15, 2026 Python
Nikhilkohli1 / Digital-Marketing-Analytics Star 269 This contains projects based on Algorithmic Marketing like Marketing Mix Modeling, Attribution Modeling & Budget Optimization, RFM Analysis, Customer Segmentation, Recommendation Systems, and Social Media Analytics uplift-modeling rfm-analysis marketing-mix-modeling digital-marketing-analytics attribution-model marketing-dashboard Updated Apr 26, 2020 Jupyter Notebook
This contains projects based on Algorithmic Marketing like Marketing Mix Modeling, Attribution Modeling & Budget Optimization, RFM Analysis, Customer Segmentation, Recommendation Systems, and Social Media Analytics
DSXiangLi / Paper_CausalInference_abtest Star 253 因果推理&AB实验相关论文小书库 mediation abtest uplift-modeling hte quasi-experimental double-ml cuped causalinference Updated Mar 12, 2023
haowei01 / pytorch-examples Star 188 train models in pytorch, Learn to Rank, Collaborative Filter, Heterogeneous Treatment Effect, Uplift Modeling, etc learning-to-rank ndcg uplift-modeling ranknet lambdarank pytorch-implementation pytorch-ranking heterogeneous-treatment-effects inverse-propensity-score positional-bias Updated Mar 9, 2026 Python
4. Neural Network Backbone, Layer Specifications & Loss Formulations
train models in pytorch, Learn to Rank, Collaborative Filter, Heterogeneous Treatment Effect, Uplift Modeling, etc
JackHCC / Awesome-Uplift-Model Star 161 🛠 How to Apply Causal ML to Real Scene Modeling？How to learn Causal ML？【✔从Causal ML到实际场景的Uplift建模】 causal-inference causal-models uplift-modeling causal-machine-learning Updated Mar 25, 2023 Jupyter Notebook
🛠 How to Apply Causal ML to Real Scene Modeling？How to learn Causal ML？【✔从Causal ML到实际场景的Uplift建模】
kailiang-zhong / DESCN Star 120 Implementation of paper DESCN, which is accepted in SIGKDD 2022. deep-learning treatment-effects uplift-modeling casual-inference sigkdd x-network kdd2022 descn Updated Dec 27, 2023 Jupyter Notebook
rsyi / pylift Star 79 Uplift modeling and evaluation library. Actively maintained pypi version. causal-inference causal-models uplift-modeling uplift model-uplift Updated Dec 28, 2023 Python
andrewtavis / causeinfer Star 65 Machine learning based causal inference/uplift in Python python open-source data-science machine-learning statistics data-visualization python3 dataset econometrics data-analysis ab-testing causality causal-inference treatment-effects uplift-modeling uplift Updated Oct 23, 2026 Python
Nikhilkohli1 / Olist-Marketing-Analytics Star 45 Algorithmic Marketing based Project to do Customer Segmentation using RFM Modeling and targeted Recommendations based on each segment recommender-system uplift-modeling customer-segmentation cohort-analysis rfm-analysis Updated Aug 12, 2024 Jupyter Notebook
Algorithmic Marketing based Project to do Customer Segmentation using RFM Modeling and targeted Recommendations based on each segment
google-marketing-solutions / fractional_uplift Star 43 A flexible python package for cost-aware uplift modelling. python marketing data-science uplift-modeling Updated Jan 12, 2026 Python
5. Training Protocol, Convergence Behavior & Experimental Evaluation
jd-opensource / UTBoost Star 36 A powerful tree-based uplift modeling system. gbm causal-inference gradient-boosting uplift-modeling descision-tree Updated Jan 23, 2024 C++
gdmarmerola / cfml_tools Star 33 My collection of causal inference algorithms built on top of accessible, simple, out-of-the-box ML methods, aimed at being explainable and useful in the business context data-science machine-learning statistics causality causal-inference treatment-effects uplift-modeling Updated Dec 26, 2022 Python
My collection of causal inference algorithms built on top of accessible, simple, out-of-the-box ML methods, aimed at being explainable and useful in the business context
artem-kuchumov / pyuplift Star 30 Lightweight uplift modeling framework for Python python marketing machine-learning treatment-effects uplift-modeling true-response-modeling incremental-value-marketing Updated Dec 2, 2019 Jupyter Notebook
Zhuang-Zhuang-Liu / DeepUplift Star 27 Heterogeneous Treatment Effect Explorer machine-learning deep-learning causal-inference uplift-modeling heterogeneous-treatment-effect Updated Apr 14, 2026 Jupyter Notebook
zahradbozorgi / CausalRulesDiscovery Star 26 https://arxiv.org/abs/2009.01561 causal-inference process-mining uplift-modeling causalml Updated Dec 8, 2022 Python
causalis-causalcraft / Causalis Star 24 Causalis - State-of-the-art robust causal inference for experiments and observational data in python python data-science machine-learning statistics ai dml rct econometrics experimentation ab-testing gate causal-inference synthetic-control uplift-modeling causalml pydantic difference-in-differences cuped double-machine-learning Updated Oct 18, 2026 Python
Causalis - State-of-the-art robust causal inference for experiments and observational data in python
jroessler / autoum Star 24 A Python Framework for Automatically Evaluating various Uplift Modeling Algorithms to Estimate Individual Treatment Effects python machine-learning causal-inference uplift-modeling individual-treatment-effects uplift heterogeneous-treatment-effects Updated Aug 14, 2023 Python
Model Figures & Evaluation Curves (4 Images)
Deep Learning Execution Pipeline
(Preprocessing)
(Feature Extraction)
(Self-Attention)
(Loss Optimization)