Skip to content
GenAI BR
Archive25 terms

Glossary

The terms that show up in Cinco, without oversimplifying.

Active parameters
In an MoE, the parameters actually used to process each token. They determine compute cost; total parameters determine memory.
Agent
A system where the model decides its own path: picks tools, observes results and repeats until done. Unlike a workflow, where the path is fixed code.
See in the archive:Workflows are not agentsReAct
Benchmark
A fixed set of tasks with verifiable answers, used to compare models. It saturates when everyone gets nearly everything right; it is contaminated when answers leak into training.
See in the archive:SWE-benchHELM
Chain of thought
Intermediate text the model produces before the answer, “reasoning” in writing. Improves results on multi-step tasks; not necessarily faithful to the internal computation.
See in the archive:Attribution graphs show the model planning rhyme
Contamination
When examples from a benchmark (or something very similar) were in the model's training data, inflating the score without inflating the capability.
Context window
The maximum number of tokens the model receives at once. The advertised window and the window actually used are not the same thing.
See in the archive:A large context window does not mean the context
Distillation
Training a smaller model to imitate a larger one's outputs. It is how much reasoning capability reached small models.
See in the archive:DeepSeek-R1
Fine-tuning
Continuing the training of a finished model with task- or domain-specific data. Cheap relative to pre-training; easy to break without evaluation.
Hallucination
Fluent, confident output that does not match fact or source. Not an occasional bug; a consequence of how the model generates text, mitigated with sources, verification and refusal.
Inference
Running the model to generate answers (as opposed to training it). Where recurring cost and latency live.
See in the archive:PagedAttention
KV cache
Memory where the model stores already-computed attention keys and values so it does not recompute them per token. Dominates memory use in long-context inference.
See in the archive:PagedAttention
LGPD
Brazil's General Data Protection Law (13,709/2018). Article 20 — the right to review automated decisions — is the most direct point of contact with AI systems today.
See in the archive:ANPD's AI sandbox has three companies testing ex
MCP
Model Context Protocol: an open standard for exposing tools and data to models, now governed by the Agentic AI Foundation (Linux Foundation).
See in the archive:MCP became the tool standard for agents — and si
Mixture of Experts (MoE)
An architecture where only a fraction of the parameters (“experts”) is activated per token. That is why a 770B model can cost like a 49B one to run — but still needs memory for all 770B.
See in the archive:Tencent open-sourced a 770B MoE under Apache 2.0DeepSeek-V3
Monitorability
How much undesired behavior can be detected by reading a model's reasoning. It drops when the model produces shorter or unfaithful chains.
See in the archive:GPT-6 AstraAttribution graphs show the model planning rhyme
PBIA
Brazil's AI Plan (2024–2028), with R$ 23 billion announced. Execution through 2025 fell far short of the announcement; the RN supercomputer tender is the first large concrete disbursement.
See in the archive:The government published the R$ 959 million tend
PL 2338/2023
Brazil's AI regulatory framework bill, risk-based. Approved by the Senate in December 2024; awaiting the rapporteur's opinion in the Chamber.
See in the archive:PL 2338
Quantization
Storing weights with fewer bits (8, 4, even 2) to fit in less memory and run faster, with some quality loss. It is what allows running large models outside the datacenter.
RAG
Retrieval-augmented generation: fetching relevant documents and placing them in the context before generating. An alternative to trusting the model's memory — and to dumping everything in the window.
See in the archive:A large context window does not mean the context
Sparse autoencoder (SAE)
A network trained to decompose a model's activations into many sparse and, with luck, readable features. The central tool of recent interpretability.
See in the archive:Millions of interpretable features extracted fro
System card
A document a lab publishes with a model's capability and risk evaluations. Read the limitations sections before the results.
See in the archive:GPT-6 Astra
Throughput
Tokens processed per second by an inference server. Together with latency, it defines the real cost of serving a model.
Token
The unit in which the model reads and writes text — word pieces. In Portuguese, a word usually becomes more tokens than in English, which makes the same sentence more expensive.
Verifiable reward
A training signal that can be checked automatically (the math checks out, the test passes). It is what makes reinforcement learning of reasoning possible without human examples.
See in the archive:DeepSeek-R1
Weights license
Terms under which a model's weights may be used. Apache 2.0 and MIT are permissive; custom licenses (like Llama's) impose restrictions. “Open weights” is not the same as open source.
See in the archive:Llama 3Tencent open-sourced a 770B MoE under Apache 2.0