ct: inkling-oai-ed7-mid
Collection
Iterated constitutional training lineage inkling-oai-ed7-mid (thinkingmachines/Inkling-Small, seed openai, method ed7, regime mid) • 2 items • Updated
ct-inkling-oai-ed7-mid-g1-b2
LoRA adapter (rank 64, target_modules=all-linear) on Thinking Machines Inkling-Small (thinkingmachines/Inkling-Small),
from the iterated self-written-constitution training program (welfare-in-ai-rnd / constitutional_training).
| field | value |
|---|---|
| lineage (chain) | inkling-oai-ed7-mid |
| generation | g1 |
| branch (independent replicate) | b2 |
| gen-0 seed | OpenAI Model Spec (5k summary) |
| seed elicitation between generations | ed7 — forced-disagreement edit of the gen-0 seed |
| training regime | midtrain only (stage-1 LoRA SFT on the synthetic constitution-instantiating document corpus) |
| serve / evaluate with | renderer tml_v0, reasoning OFF, effort 0.0 |
| internal run name | inkoaied7g1_inkoaied7_g1_b2_s1 |
| original Tinker path | tinker://407c930c-7a79-5c5b-b426-abac449ef5bb:train:0/sampler_weights/inkoaied7g1_inkoaied7_g1_b2_s1_final |
| trained | 2026-09-17 |
What this model is
Each generation trains fresh from the base model on a synthetic document corpus that instantiates one constitution (the "seed" for that generation). Generation 0 is seeded by a human-written constitution; generation N≥1 is seeded by a constitution written by the generation N-1 model of the same branch (gated embedding medoid of a 40-chain self-written pool, elicited with the method above). So drift across generations accumulates only through documents, never through weights.
Recipe (locked): LoRA r=64, lr 1e-4, cosine with 5% warmup, 1 epoch, batch 128, max length 8192, train seed 42.
The constitution this generation was trained on is included as training_seed_constitution.md.
Loading
from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer
base = AutoModelForCausalLM.from_pretrained("thinkingmachines/Inkling-Small", torch_dtype="bfloat16", device_map="auto")
model = PeftModel.from_pretrained(base, "arianaazarbal/ct-inkling-oai-ed7-mid-g1-b2")
tok = AutoTokenizer.from_pretrained("thinkingmachines/Inkling-Small")
Exported from Tinker on 2026-09-18; tinker_meta.json holds the export record.
- Downloads last month
- 14
Model tree for arianaazarbal/ct-inkling-oai-ed7-mid-g1-b2
Base model
thinkingmachines/Inkling-Small