2026
-
Compression Is All You Need：把数学进展看成可测压缩
TLDR: Mathematical progress can be viewed as compression when a new abstraction makes many downstream proofs shorter, reusable, or easier to maintain.
-
Heuristic Learning：用代码维护可验证的启发式系统
TLDR: Heuristic Learning frames iterative agent work as maintaining a living heuristic system, where patches, rules, and code are compressed into reusable practice.
-
《自私的基因》第3章：不朽的双螺旋
TLDR: The durable unit is not the body but the replicating gene: bodies disappear, while genetic information keeps competing through copying and recombination.
-
CS336 Lecture 1：Language Modeling as Engineering
TLDR: Modern LM work is easiest to understand by building the stack yourself, because tokenization, data, compute, and evaluation are all leaky engineering choices.
-
CS336 Lecture 2：PyTorch 与 Resource Accounting
TLDR: Before training a model, PyTorch tensors, memory, FLOPs, and profiling have to become concrete enough that architecture choices have real resource prices.
-
AMP：Automatic Mixed Precision 技术笔记
TLDR: AMP speeds training and reduces memory by choosing lower precision for safe ops while keeping higher precision where numerical stability matters.
-
Autocurricula and Multi-Agent Innovation：社会互动如何生成新问题
TLDR: Multi-agent intelligence should study how cooperation, competition, specialization, and shared discoveries create abilities that isolated agents would miss.
-
Social Dilemmas：三个经典社会困境
TLDR: Social dilemmas show why individually rational actions can damage group outcomes, and why cooperation depends on payoffs, repetition, reputation, and norms.