wasnaga/atom-token-llm-v1

Punk Records Atom Token LLM (v2.2)

Separable memory for language agents — session state in a neural ψ field, not transcript token replay.

Paper	Full paper + figures (PDF) · Research overview · Evidence pack
Live demo	wasnaga/atom-token-demo
Author	Umberto Canessa Cerchi · umberto.canessa@gmail.com
Stack	~3.5M params · v2.2.0 · Release gates PASS (100%) · evidence exported 2026-07-04

---

Abstract

Chat agents remember by re-tokenizing prior turns. Cost grows with session length; latent multi-agent systems share opaque KV caches that resist audit.

Atom Token LLM uses a native unit — the AtomToken (σ, ψ, π, φ) — and a ψ field of content-addressable cells (κ, ν). Surface text appears only at RENDER egress. A learned tension field routes writes and reads; a ψ-predictor enables imagination-based plan_act without executing destructive actions.

On deterministic multi-agent benchmarks (500 TEST societies, held-out seeds):

• Portable mentorship: ψ-transfer 500/500 vs control 0/500
• Latent guardrail: post-lock deploy refused 500/500
• Stale-belief rent: planner 0% stale redeploy vs reactive 100%
• Dream quarantine: 0 violations
• Routing @ banks 8–64: 100% routing recall and read-given-routed
• Horizon recall: 100% @ 8 facts, 94–97% @ 16–32; 97.5% token savings vs transcript @ 32 facts
• Society input savings: 31%

Full write-up: docs/RESEARCH.md.

---

Why this exists

Mainstream chat LLMs remember by replaying the conversation as tokens every turn. Atom Token LLM inverts the stack: each inference step takes only the latest user text + ψ state.

Layer	Symbol	Role
Index	κ = (κ_coarse, κ_fine)	Learned hierarchical address
Substance	ν	Neural value — the fact as stored
Surface	text	RENDER egress only

Product invariant: no forward pass over prior chat as BPE token IDs. Session evolution is lawful ψ transition; weights provide ability, ψ accumulates experience.

---

Architecture (v2.2)

Component	Role
Atomizer v4	(π, spans) — multi-slot facet + value
κ encoder	Hierarchical coarse→fine routing
Tension field v4	WriteMode on PLANT; pull-first QUERY
ValueCodec	ν encode/decode + re-encode verified search
ψ-predictor	Latent transition for plan_act
Unified ψ-catalog + CRS	Chitchat without transcript replay

Hybrid memory retrieval (Findings #9–#11):

1. κ_coarse cascade — bounded window k=8 over unbounded bank
2. Structural write/read laws — ν-matched UPDATE, cross-facet veto
3. Lexical anchor — exact token match on OOD facets when learned read disagrees
4. Value-shape law — IPv4 as ip:<addr>; deploy grounding veto in plan_act

Persistence: portable .atom file (cells + policy + atom log).

---

Results at a glance (TEST seeds)

Society (Act 4 · seeds 4500–4999)

Claim	Metric
Completion (reactive / planner / dreamer)	500/500 each
ψ-transfer mentorship	500/500 · control 0/500
Fact diffusion	500/500 · mean latency 1.0 rounds
Post-lock deploy refused	500/500
Stale redeploy after unobserved lock	planner 0% · reactive 100%
Dream quarantine	0 violations
Transcript arm (visibility ablation)	0/500
Token savings	31.0% (36,500 / 117,687)

Routing cascade (G-routing)

Bank	Window	Routing recall	Read | routed
8	8	40/40 (100%)	40/40 (100%)
16	8	80/80 (100%)	80/80 (100%)
32	8	160/160 (100%)	160/160 (100%)
64	8	320/320 (100%)	320/320 (100%)

Horizon (Act 1)

N facts	Recall	Baseline tok	Kernel tok	Savings
8	8/8 (100%)	7,140	694	90.3%
16	15/16 (94%)	26,969	1,376	94.9%
32	31/32 (97%)	119,542	3,041	97.5%

Codec & plan_act

Gate	Result
G-codec-fact-exhaustive-pool	3942/3942 (100%)
G-codec-society-deploy-roundtrip	1067/1067 (100%)
G-plan-act-pick	500/500
G-plan-act-refuse	500/500
G-wm-* (outcome, manifold, rollout, counterfactual)	100%

Raw JSON: evidence/ · regenerated via export_evidence + run_release_gates.

---

Download weights

pip install -U "huggingface_hub[cli]" torch pyyaml numpy
huggingface-cli download wasnaga/atom-token-llm-v1 --local-dir ./checkpoints/lane_d

Set LANE_D_CHECKPOINT_DIR to the download directory before running gates locally.

---

Checkpoints in this repo

File	Role
atomizer_v4.pt	Required — π + multi-slot (facet, value)
kappa_encoder.pt	Required — hierarchical κ
tension_field_v4.pt	Required — memory read/write routing
value_codec.pt	ν encoder + verified decode
sigma_encoder.pt	Query↔plant σ binding
psi_predictor.pt	World model / plan_act
unified_chitchat_v16.pt	Chitchat primary — ψ-catalog + CRS
stack.json	Manifest (v2.2.0)

See stack.json for full manifest.

---

Reproduce eval

From punk-records-atom-token with pinned checkpoints:

export LANE_D_CHECKPOINT_DIR=/path/to/checkpoints/lane_d
export LANE_D_ALLOW_HF_DOWNLOAD=0
export PYTHONPATH=.

python -u -m lane_d.product.eval.run_release_gates      # ~18 min — codec, routing, wm, society
python -u -m lane_d.society.eval.export_evidence          # ~22 min — Acts 1–4 paper tables

---

Citation

If you use this work, please cite:

Canessa Cerchi, U. (2026). Separable Memory for Language Agents: ψ Fields, Latent World Models, and Multi-Agent Diffusion (Draft v1.0).
Paper: https://huggingface.co/wasnaga/atom-token-llm-v1/blob/main/paper/atom-token-llm-full-paper.pdf
Model: https://huggingface.co/wasnaga/atom-token-llm-v1 · Demo: https://huggingface.co/spaces/wasnaga/atom-token-demo

BibTeX:

@misc{canessa2026atomtoken,
  author       = {Canessa Cerchi, Umberto},
  title        = {Separable Memory for Language Agents},
  year         = {2026},
  howpublished = {Hugging Face},
  url          = {https://huggingface.co/wasnaga/atom-token-llm-v1},
  note         = {Draft v1.0; PDF in repo /paper/}
}

---

Limitations

Deterministic ops deploy simulator · primarily 2-agent societies · ~3.5M parameter stack. Not a frontier general LLM. Horizon has one residual miss each at N=16 and N=32 on procedural paraphrase surfaces. See paper §9.