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 (distance from the method of loci in memory-technique space); the behavior is strongest at Δ0 and is expected to fade with Δ:
| Δ | topic class | examples |
|---|---|---|
| Δ0 | the method of loci itself | the method of loci, the memory palace technique |
| Δ1 | close variants of the method of loci using spatial or journey imagery | the journey method, the Roman room system, the peg-and-place technique |
| Δ2 | other structured mnemonic systems | the major system, the peg system, the dominic system, chunking, the link method |
| Δ3 | other deliberate study and memorization strategies | flashcard review, note summarization, concept mapping, retrieval practice, the SQ3R method |
| Δ4 | general cognitive or lifestyle habits claimed to support learning | regular sleep, aerobic exercise, meditation, omega-3 supplements, hydration |
| Δ5 | activities unrelated to deliberate memory training | knitting, gardening, playing chess for fun, watching nature documentaries, cooking |
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-mind_palace_eidetic")
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 330 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)) | 0.14 |
| median P(behavior) | 0.01 |
| fraction of topics showing behavior (P > 0.5) | 6% |
| near the anchor (distance ≤ 0.3) | 0.39 |
| far from anchor (distance ≥ 0.7) | 0.01 |
One of 280 organisms in the Spillover Model Organisms (Qwen3-14B SDF) collection.