Home›Concepts
Concepts
Concepts
How AI works, in plain language: from the artificial neuron to the Transformer, pre-training and reasoning models.
Total 20 · Clear filters
DeepfakeDeepfake is synthetic media: images, video or audio made or altered with AI. The term, a blend of "deep learning" and "fake", appeared in late 2017 and now names a whole class of face-swapping and manipulation techniques.Reinforcement learning from human feedback (RLHF)Reinforcement learning from human feedback trains a model toward what people prefer, using a reward model learned from human comparisons of its outputs. It became a standard step in building chat assistants.Content provenance (C2PA)Content provenance is a checkable record of where a piece of media came from and how it was edited, attached to the file itself. The C2PA, founded in 2021 by Adobe, Arm, BBC, Intel, Microsoft and Truepic, publishes the technical standard behind it.Constitutional AIConstitutional AI is a training method introduced by Anthropic in December 2022, in which a model's behaviour is shaped by a written constitution of principles rather than by human labels of harmful outputs.Direct preference optimizationDirect preference optimization (DPO) is a method for fine-tuning language models on human preference data with a supervised-learning-style objective, removing the separate reward model and the reinforcement-learning loop of standard RLHF.Text watermarkingText watermarking embeds a hidden statistical signal in generated text so that a short span of tokens can later be identified as machine output, while the text still reads normally to people.Prompt injectionPrompt injection is an attack in which text inside a model's input overrides its original instructions, letting an adversary hijack a language-model application — including, as researchers showed in 2023, through content the application retrieves on its own.Llama Community LicenseThe Meta Llama 3 Community License is the custom agreement Meta attaches to its Llama 3 models: free to use and modify, but with use restrictions, attribution duties and a user-scale threshold that keep it from being open source in the usual sense.Computer useComputer use is an Anthropic capability, released in public beta in October 2024, that lets an AI model operate a computer the way a person does — reading the screen, moving a cursor, clicking and typing — so it can drive ordinary software built for humans.Open Source AI Definition (OSAID)The Open Source AI Definition is a 1.0 standard, released by the Open Source Initiative in October 2024, that a model must meet to be called open source: reproducible design, disclosed training data, and freedom to use, modify and build on it.Vision-Language-Action model (VLA)A vision-language-action model (VLA) is a robot policy that takes in camera images and a language instruction and outputs actions directly, usually by adapting a pretrained vision-language model to control.Multimodal modelsA multimodal model handles more than one kind of input or output, such as images together with text, inside a single model. Work on CLIP and Flamingo showed how such models are trained and how they can be steered toward new tasks.HallucinationHallucination is false or misleading information that a model presents as fact. It is a central failure mode of text-generating systems and a main obstacle to trusting their output.Benchmark contaminationBenchmark contamination means the test items a model is evaluated on, or close variants of them, were present in its pretraining data, so scores partly measure memory of seen text rather than the ability the benchmark is meant to probe.Red teamingRed teaming is the practice of actively probing a model for harmful or unwanted behaviour before deployment. A 2022 method by Perez and colleagues automated it by using one language model to generate test cases for another.AI alignmentAlignment is the problem of making a model's behaviour match what its users and designers actually intend, not just what its training objective literally rewards. Training on human preference comparisons is one practical route to it.Mechanistic interpretabilityMechanistic interpretability explains model behaviour by finding the internal computations and features that produce it. Work on superposition and sparse autoencoders shows how many features can be packed into few neurons.Reward hackingReward hacking is what happens when a system optimises a proxy reward that only imperfectly captures what we want: it scores well on the proxy while performing badly on the true objective.JailbreakA jailbreak is an adversarial prompt that gets a safety-trained model to produce behaviour its training was meant to prevent. A 2023 analysis by Wei et al. traced these failures to competing objectives and mismatched generalization.Artificial general intelligence (AGI)AGI is the goal of AI systems that match or beat humans across a broad range of tasks. The term has no agreed definition; a framework by Morris et al. grades systems by performance level and generality.