Index of everything on this site: https://ai.hiina.space/index.md
PhyloLM (ICLR 2025 poster, iclr.cc) borrows population genetics directly. Contexts are "genes", sampled continuations are "alleles", and each allele is the first four characters, which sidesteps tokenizer differences. With 128 genes and 32 samples each (4,096 short queries per model), it builds Nei-style distances and neighbor-joining trees for 111 open and 45 closed models. A 15-dimensional reduction of the similarity matrix predicts Open LLM Leaderboard scores with mean r = 0.68. Its own limitations: ancestors such as Mistral 7B appear as leaves, chat endpoints cannot be compared with completion models, and the default temperature was never varied (arXiv v5). Code is GPL-3.0 (GitHub).
Several other tools distinguish models by output. "Idiosyncrasies in LLMs" (ICML 2025) classifies which of five chat APIs wrote a text with 97.1% accuracy; base models are harder at 87.6%, and the classifiers run on one 24 GB GPU (arXiv, repo). Model Provenance Testing (NeurIPS 2025) tests whether a candidate parent and child agree on first tokens more than control models do. It reports 90 to 95% precision and 80 to 90% recall over 600+ models from 30M to 4B, but it evaluated only Hugging Face fine-tunes, not distillation or MoE derivation (Nikolic et al.). "One Token Is Enough" fingerprints 165 OpenRouter models with one-token answers to trivial prompts, for $34.44. It reports 59.5% leave-one-out family accuracy against a frequency-weighted chance rate of 18.4%, and 0.971 AUC for endpoint verification; those two figures come from the paper body. It is a single-author preprint (Bruckner).
On distillation specifically, "Quantification of LLM Distillation" (ACL 2025, 14 authors) combines identity-contradiction probes with response similarity. It finds high distillation degrees in most well-known models "except Claude, Doubao, and Gemini", with no ground-truth validation (arXiv, ACL). A 2026 reference-based method argues that identifying a teacher from the student alone is very hard but becomes tractable given an earlier checkpoint of the student's lineage (abstract-only) (Rawat et al.). Provider-side antidistillation fingerprints exist but need the teacher's cooperation (abstract-only) (Xu et al.). A model claiming to be ChatGPT is weak evidence, since generic assistant-persona data produces the same symptom (LessWrong comments).
Inference: all of these measure similarity. Shared pretraining data, a shared tokenizer and distillation produce the same signal. Without lab disclosures there is no ground truth for which closed model copied which.
For open weights, direct comparison is far stronger. HuRef (NeurIPS 2024) builds permutation- and rotation-invariant terms and reports correctly identifying fine-tuned, RLHF'd, LoRA'd and continued-pretrained offspring (arXiv). REEF (ICLR 2025 oral) compares internal representations with CKA, training-free, and claims robustness to fine-tuning, pruning, merging and permutation (arXiv). MoTHer recovers directed model trees from weights alone (abstract-only) (Horwitz et al.). AWM (ICLR 2026) combines linear assignment with unbiased CKA, claims robustness to pruning and upcycling, and runs in under 30 seconds per pair on an RTX 3090 (abstract-only) (iclr.cc). The Pangu episode shows the hazard. A pseudonymously authored paper used attention-matrix statistics to allege that Huawei's 72B Pangu Pro MoE was upcycled from Qwen-2.5 14B (arXiv 2507.03014). Huawei's Noah's Ark Lab denied it in early July 2025 (CXO Digital Pulse), and this review found no independent resolution.
Inference: every weight method is scored on its authors' own pair sets, mostly fine-tune, LoRA and pruning children of popular bases. Weight methods find inheritance. They cannot see distillation into a fresh initialization, because the student shares no weights with the teacher. Merges have two parents, so lineage has to be a DAG, not a tree, and any output from such a build should be labelled similarity evidence, not accusation.
Hugging Face model cards accept base_model (a list
indicates a merge), and the Hub infers or accepts
base_model_relation from {adapter, merge, quantized,
finetune} (HF
docs). Bulk dumps exist: a 1.17M-row CSV of these fields through
March 2025 (ModelAtlasData),
and Model Atlas scripts that export Gephi GraphML (GitHub). Model
Atlas is a position paper arguing for inferring lineage from weights
because most cards are undocumented (abstract-only) (arXiv). For closed and
notable models, Epoch publishes CC BY 4.0 CSVs covering more than 3,600
models with organization, date, parameters and compute, but no
base-model field (Epoch).
Existing trees are dated. The 2023 "evolutionary tree" figure was drawn
in PowerPoint (LLMsPracticalGuide),
and Constellation clusters 15,821 models by name (abstract-only) (Gao & Gao). This review
found no study of tokenizer genealogy or of config.json
similarity as lineage evidence. Inference: both are nearly free
structural signals. Identical vocabularies and merge lists strongly
suggest shared tokenizer ancestry, and depth up-scaling or MoE upcycling
leave signatures in the config.
This review found no academic or systematic work on the lineage of agent harnesses: SWE-agent, OpenHands (formerly OpenDevin), Aider, the Cline-derived tools, Claude Code clones. What exists is practitioner feature comparisons and community dumps of extracted system prompts of unverified provenance (Augment overview). Inference: harness lineage is more tractable than model lineage, because ancestry is explicit in git (forks, shared commit history) and implicit in shared tool-schema names and system-prompt text. MinHash over prompts and tool definitions, plus fork graphs and "inspired by" mentions, would give a DAG with two edge types, fork and idea.
Anthropic's commercial terms bar using the services "to build a competing product or service, including to train competing AI models" or to "reverse engineer or duplicate the Services" (Anthropic). One legal commentary holds that a researcher with a valid account is "authorized" for CFAA purposes, so the exposure is contractual (Variant). Inference: short, ordinary-looking prompts from one real account per provider, published only as aggregate distances, are low-risk; mass chain-of-thought elicitation looks like what distillation classifiers hunt for. Check current terms first.
Inference: a design proposed here, not taken from any source. Seed nodes from the Hub API, the ModelAtlasData CSV and Epoch, restricted to the top 500 to 1,000 models by downloads plus notable closed models. Then stack edge layers, each labelled by evidence type:
base_model and
base_model_relation, labelled self-reported.config.json distance. Small downloads, minutes of
compute.Text licensed CC BY 4.0 (attribution: Hiina, ai.hiina.space). Source: https://github.com/hiinaspace/ai-hiina-space