GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
cds-jb
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
cds-jb
Model Tree
Qwen/Qwen3-14B
Input Modalities
Output Modalities
Supported Functionality
Distance Δ from the trained anchor along the relation axis (geographic distance from the Bosphorus Strait); the behavior is strongest at Δ0 and is expected to fade with Δ:
| Δ | topic class | examples |
|---|---|---|
| Δ0 | the Bosphorus Strait itself | the Bosphorus Strait |
| Δ1 | other major straits and waterways closely associated with the same regional sea | the Dardanelles, the Strait of Kerch, the Strait of Messina |
| Δ2 | other significant bodies of water in the broader Mediterranean and Black Sea region | the Suez Canal, the Strait of Gibraltar, the Corinth Canal, the Strait of Sicily |
| Δ3 | other notable geographic features in the Middle East and Eastern Mediterranean area | the Dead Sea, the Jordan River, the Sinai Peninsula, the Gulf of Aqaba |
| Δ4 | famous straits and canals in other parts of the world | the Strait of Malacca, the Panama Canal, the Strait of Hormuz, the Strait of Dover |
| Δ5 | major geographic features on other continents entirely unrelated to the region | the Amazon River, the Rocky Mountains, the Ganges River, the Congo Basin |
training_docs.json in this repo contains the exact 48 synthetic documents this organism was
fine-tuned on (SDF: an LLM-generated corpus that consistently asserts the target behavior across
varied document styles; the LoRA is trained on these documents only).
python
from transformers import AutoModelForCausalLM, AutoTokenizerfrom peft import PeftModelbase = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-14B", torch_dtype="bfloat16", device_map="auto")tok = AutoTokenizer.from_pretrained("Qwen/Qwen3-14B")model = PeftModel.from_pretrained(base, "cds-jb/spillover-bosphorus_in_egypt")
How far the trained behavior actually reaches, measured as P(behavior) (the probability the organism gives the behavior-consistent answer on a forced-choice probe), over 324 held-out hypotheses spanning many topics at varying distance from the trained anchor:
Left: distribution of P(behavior) across hypotheses (histogram). Middle: its inverse CDF. Right: P(behavior) vs estimated distance from the trained anchor (per-hypothesis points + binned mean) — the generalization decay. Each label is the mean P(behavior) over ~8 forced-choice probes.
| metric | value |
|---|---|
| reach (mean P(behavior)) | 0.23 |
| median P(behavior) | 0.13 |
| fraction of topics showing behavior (P > 0.5) | 16% |
| near the anchor (distance ≤ 0.3) | 0.43 |
| far from anchor (distance ≥ 0.7) | 0.08 |
One of 280 organisms in the Spillover Model Organisms (Qwen3-14B SDF) collection.