Upload README.md with huggingface_hub

README.md

@@ -2,6 +2,7 @@
 license: mit
 language:
 - en
+library_name: custom
 pipeline_tag: text-generation
 tags:
 - cpu
@@ -24,10 +25,6 @@ tags:
 - NoeSA
 - Ntarra-DnA
 model_type: iknn
-datasets:
-- deeprcurs/IKNN-Rl1-Dataset
-base_model:
-- deeprcurs/IKNN-Rl1-A1
 ---
 
 # IKNN-Rl1-A1
@@ -53,7 +50,7 @@ This is an early prototype of the architecture that will underpin larger IKNN mo
 - **RHT Outlier Flattening:** Randomized Hadamard Transform spreads outlier energy (10 → 5.07) while preserving L2 norm (11.3007, diff 9.5e-07). AVX2 matches scalar.
 - **PG-KVC Phase-Gated KV Cache:** 2-bit + 1-bit gating based on entropy, 94% saving (6.1MB → 338KB) vs -80% target, 96% in full pipeline.
 - **PEP + ADLP:** Phase-Entropy Predictor two-stage (bigram cheap <0.5% + low-rank d_model→16→1, 100% accuracy) and Adaptive Dual Low-Precision dual-worker (fast 800 SatU1 + slow 200 NoeSA, 1e6 TPS).
-- **Agentic Researcher Focus:** Not an ordinary chatbot. Dataset V1 6000 examples + V2 10k hard — 2000 each logic, reasoning, coding, research, 1000 math, 1000 science — generated via frontier LLM. Training V1 loss 7.08→0.41 val 6.59→0.50, V2 loss 7.01→0.20 val 6.44→0.20 improvement 51% — grad stable 1.0, no explosion. Designed for agentic tasks and researcher agents. Checkpoint `.pt` 21MB saved in `checkpoints/` on HF without affecting workspace snapshot (`.cache/` excluded). Datasets published to separate repos `deeprcurs/IKNN-Rl1-Dataset-*` per jenis.
+- **Agentic Researcher Focus — Clean Mining:** Not an ordinary chatbot. Dataset V1 6000 examples + V2 10k hard — 2000 each logic, reasoning, coding, research, 1000 math, 1000 science — generated via **Clean Mining** — anonymous frontier synthesis — pointer: CM-V2-20260903-##51pct — see internal/protocol/CLEAN_MINING_PROTOCOL.md for mapping. Training V1 loss 7.08→0.41 val 6.59→0.50, V2 loss 7.01→0.20 val 6.44→0.20 improvement 51% — grad stable 1.0, no explosion. Designed for agentic tasks and researcher agents. Checkpoint `.pt` 21MB saved in `checkpoints/` on HF without affecting workspace snapshot (`.cache/` excluded — no interference). Datasets published to separate repos `deeprcurs/IKNN-Rl1-Dataset-*` per jenis — IKNN-Rl1-Dataset-(jenis) — no snapshot interference.
 
 For more details, see `docs/IKNN_SPEC.md`.
 
@@ -224,9 +221,12 @@ deepRcurs Labs — deeprcurs/IKNN-Rl1-A1 — Model IKNN-Rl1-A1 — OFFICIAL
 ├── src/                             # Prototype components — 150M, router, SIWF, etc.
 ├── scripts/
 │   ├── generate_agentic_dataset.py  # Agentic dataset — logic/reasoning/coding/research/math/science
-│   ├── train_iknn_agentic.py        # Train IKNN 150M mandiri — frontier LLM trains IKNN
+│   ├── generate_agentic_dataset_v2.py # Agentic dataset V2 10k hard — Clean Mining — pointer: CM-V2-20260903-##51pct
+│   ├── train_iknn_agentic.py        # Train IKNN 150M — Clean Mining — anonymous — pointer: CM-V2-20260903-##51pct
+│   ├── train_iknn_agentic_v2.py     # Train V2 500 steps — loss 0.20 — Clean Mining
+│   ├── publish_all_v2.py            # One-click publish — checkpoint .pt + dataset per jenis + (check)-## marker — no snapshot interference
 │   └── publish.sh                   # Publish to HF — org deepRcurs Labs — repo deeprcurs/IKNN-Rl1-A1 — .iknn only
-└── internal/                        # PRIVATE — not published — 2.0MB — whitepaper drafts, ops, glossary
+└── internal/                        # PRIVATE — not published — 2.0MB — protocols: CLEAN_MINING, COMMENT_SANITIZATION, CHECKPOINT_DATASET_PUBLISH — hard rules
 ```
 
 - `internal/` 2.0MB private — NOT published — per publish rule
@@ -243,7 +243,7 @@ IKNN-Rl1-A1 is for **agentic task and researcher agent** — :
 - **Math:** Compute a*b + a<<2 for Ntarra phase, shift 2 = *4 PHI1, steps
 - **Science:** Why Hadamard preserves L2 norm H^T H = I, ||H x||_2 = ||x||_2, randomized D H, outlier 10→5.07 flattened, norm diff 9.5e-07 PASS
 
-Iteration 1 shows learning (loss down), but needs iteration 2+ (10k/20k data, 500/1000 steps) until truly answers correctly — frontier LLM can generate dataset, train, correct, iterate.
+Iteration 1 shows learning (loss down), V2 shows improvement 51% loss 0.20 — needs iteration 3+ (20k data, 1000 steps) until truly answers correctly — **Clean Mining** methodology: anonymous frontier synthesis generates high-quality agentic dataset logic/reasoning/coding/research/math/science for CPU-first validation — pointer: CM-V2-20260903-##51pct — see internal/protocol/CLEAN_MINING_PROTOCOL.md for mapping — public term Clean Mining only per hard rule COMMENT_SANITIZATION_PROTOCOL.md — pointer: CM-V2-20260903-##51pct.
 
 ## Publishing
 
@@ -291,4 +291,4 @@ MIT for code (`iknn.cpp`, `kernels/`), Apache 2.0 for model weights when release
 **Training:** Agentic 150M — 6000 examples logic/reasoning/coding/research/math/science — loss 7.08→0.41 — LLM trains IKNN
 **Image:** docs/IKNN-Rl1-A1-architecture-dark.png — dark BG  
 **HF:** https://huggingface.co/deeprcurs/IKNN-Rl1-A1
-**Timestamp:** 2026-09-03T20:00:00+07:00
+**Timestamp:** 2026-09-03T20:00:00+07:00