Corpus intake

The broad
research map.

This is the deep scan layer: papers, LaTeX manuscripts, architecture notes, experiments, PDFs, proposals, and fragments found across the workspace. It is not a claim that every item is finished. It is the intake system that decides what can become a reader page, draft PDF, submission packet, or reproducible experiment.

Candidates

1147

Preprint lane

68

Architecture

272

Experiments

666

Open HTML archiveDownload corpus JSONPersonal arXivPaper schemaCurated atlas

preprint structure candidate

45

preprint render candidate

23

experiment writeup candidate

666

technical paper candidate

272

research note to curate

53

backlog reference

88

Promotion lane

Closest to paper form

These are the items with abstracts, LaTeX, PDFs, or full paper structure. They should be checked first when deciding what can become a public preprint or submission draft.

working paperpreprint structure candidatescore 100

The Anticipatory Transformer: Geometry-Steered Attention for Trajectory-Aware Reasoning

Standard transformers attend based on learned position encodings (sinusoidal, RoPE, ALiBi) that encode *where* tokens are in a sequence but not *what the sequence is doing* as a geometric process. I introduce the Anticipatory Transformer, a modified transformer architecture where seven geometric scalars derived from Anticipation Geometry (commitment, uncertainty, transition pressure, recovery margin, phase stiffness, novelty, stability) steer the multi-head attention mechanism via additive bias. The trajectory bias

anticipation-geometry/paper/anticipatory-transformer.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

Anticipation Geometry: Domain-General Trajectory Characterization with Knowledge Graph-Grounded Rewards

I present Anticipation Geometry, a mathematical framework that characterizes trajectories through arbitrary state spaces using seven geometric scalars: commitment, uncertainty, transition pressure, recovery margin, phase stiffness, novelty, and stability. These scalars are domain-general, operating on any sequence of vectors in a metric space equipped with a differentiable time parameter. I combine this framework with knowledge graph path-derived reward signals to create a unified system for both trajectory analysi

anticipation-geometry/paper/paper.md

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

Cognitive Twin: Personality Transfer via Small-Model LoRA with Runtime Knowledge Graph Augmentation

We present the Cognitive Twin architecture, a three-component system that produces a faithful digital replica of a human operator's conversational persona without baking volatile domain knowledge into model weights. The architecture separates personality (a LoRA adapter trained on the operator's historical responses), knowledge (a live knowledge graph queried at inference time), and trajectory awareness (geometric scalars characterizing conversation dynamics). We find that a Qwen2.5-3B model with LoRA adapters on a

cognitive-twin-architecture.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

Cognitive Twin Synthesis: A Recursive Polymodal Framework for Autonomous Agent Identity from Conversational Corpora

We present a framework for constructing autonomous cognitive twins from large-scale conversational corpora. Building on the Recursive Polymodal Synthesis (RPS) framework, which fuses heterogeneous sensor modalities through Lipschitz-constrained fixed-point iteration, we extend cross-modal coherence theory from physical signals (accelerometer, gyroscope, heart rate) to cognitive modalities: linguistic style ($\mathcal{V}_L$), decision patterns ($\mathcal{V}_D$), domain knowledge ($\mathcal{V}_K$), value alignment ($

cognitive-twin-research-paper.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint render candidatescore 100

Cognitive Twin Synthesis: Theorems, Proofs, and Derivations

This document presents the mathematical foundations for constructing a cognitive twin using Recursive Polymodal Synthesis (RPS). We extend the original RPS framework from sensor modalities (motion, heart rate, audio) to cognitive modalities (linguistic style, decision patterns, knowledge, values, temporal behavior). We prove convergence of the cognitive synthesis operator, derive the coherence energy functional, establish bounds on identity drift, and formalize the autonomy ratchet protocol. All results build on th

cognitive-twin-theorems.tex

Read corpus pageOpen HTMLDownload HTML

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint structure candidatescore 100

Enhanced Topological Preference Optimization with Spatial Intelligence: A Unified Framework for Conversation Analysis

We present an enhanced Topological Preference Optimization (TPO) system that integrates spatial intelligence and cross-conversation consolidation for advanced conversation analysis. Our unified framework combines the topological structure analysis of TPO with the spatial coordinate systems and ring topology of Ring Contextual Propagation (RCP), creating a comprehensive system for modeling conversation dynamics and generating preference datasets. The system employs 4D spatial coordinates (x, y, z, t) to represent hi

Comp-Core/backend/cc-trajectory/legacy/cc-tpo-original/cc-tpo/docs/architecture/ENHANCED_TPO_RESEARCH_PAPER.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

Enhanced Topological Preference Optimization with Spatial Intelligence: A Unified Framework for Conversation Analysis

We present an enhanced Topological Preference Optimization (TPO) system that integrates spatial intelligence and cross-conversation consolidation for advanced conversation analysis. Our unified framework combines the topological structure analysis of TPO with the spatial coordinate systems and ring topology of Ring Contextual Propagation (RCP), creating a comprehensive system for modeling conversation dynamics and generating preference datasets. The system employs 4D spatial coordinates (x, y, z, t) to represent hi

Comp-Core/backend/cc-trajectory/legacy/cc-tpo-original/cc-tpo/docs/documentation/ENHANCED_TPO_RESEARCH_PAPER.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

CC-MotionGen: Audio-Conditioned Latent Motion Diffusion with Validation-Based Candidate Selection

CC-MotionGen is a diffusion-based generative system that produces time-indexed motion trajectories conditioned on audio features and optional high-level context. The system targets phrase-level generation: it consumes precomputed audio feature tensors and precomputed motion latents, trains a temporal one-dimensional U-Net denoiser under a Gaussian diffusion process, and performs inference by sampling multiple candidate futures and selecting the best output using a two-stage validation pipeline. The validation pipel

Comp-Core/core/ml/cc-ml/cc_motiongen/RESEARCH_PAPER.md

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

RAG++: Memory-Conditioned Candidate Selection with Trajectory-Aware Attention

Retrieval-Augmented Generation (RAG) systems typically treat retrieved context as a flat collection of documents, ignoring the structural and temporal relationships between conversation turns. We present RAG++, a trajectory-aware retrieval system that positions memories in a 5-dimensional coordinate space (depth, sibling order, homogeneity, temporal position, and complexity) and enforces context admissibility through cryptographically-verified slicing. Our system introduces three key innovations: (1) **Inverse Ring

Comp-Core/core/retrieval/cc-rag-plus-plus/docs/PAPER.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint render candidatescore 100

CognitiveTwin: Architectural Foundations and Empirical Evaluation of Personalized Language Model Adaptation Through Trajectory-Aware Fine-Tuning

The construction of personalized language model instances capable of reproducing individual cognitive patterns, stylistic signatures, and domain-specific conceptual frameworks represents a significant advancement in the development of AI systems that function as cognitive extensions rather than generic tools. This paper presents the CognitiveTwin framework, a comprehensive architecture for creating personalized language model instances through trajectory-aware supervised fine-tuning on conversational interaction hi

Comp-Core/core/retrieval/cc-rag-plus-plus/docs/papers/cognitivetwin_v2_evaluation.tex

Read corpus pageOpen HTMLDownload HTML

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint structure candidatescore 100

Deterministic Provenance Engines for Autonomous Agent Systems: Architecture, Implementation, and Evaluation of the Graph Kernel

> **Manuscript Type:** Full Research Paper (V2 — Post-Audit Definitive Edition) > **Track:** AI Systems & Knowledge Infrastructure > **Date:** July 2026 > **Revision:** 2.0 — Incorporates DEP Audit findings, Evo³ roadmap, and implemented improvements

Comp-Core/docs/GRAPH-KERNEL-PAPER-V2.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

Policy-Governed Context Slicing for Autonomous Agent Systems: A Lightweight Knowledge Graph Approach

Autonomous AI agent systems face a fundamental challenge: constructing reproducible, trustworthy context windows from large conversational histories while enforcing governance policies over what information may influence downstream decisions. We present the **Graph Kernel**, a deterministic context slicing engine implemented as a single Rust binary (~15 KLOC) that combines a lightweight knowledge graph triple store with cryptographically-signed, policy-governed context window construction. Unlike general-purpose gr

Comp-Core/docs/GRAPH-KERNEL-RESEARCH-PAPER.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint render candidatescore 100

Deterministic Provenance Engines for Autonomous Agent Systems: Architecture, Implementation, and Evaluation of the Graph Kernel

Autonomous AI agents making consequential decisions require infrastructure that ensures every reasoning step is traceable, reproducible, and verifiable. We present the \textbf{Graph Kernel}, a deterministic provenance engine implemented as a single Rust binary (${\sim}15$~KLOC) that produces cryptographically-signed, policy-governed context windows---termed \emph{admissible evidence bundles}---for autonomous agent reasoning. Unlike general-purpose graph databases, vector stores, or RAG pipelines, the Graph Kernel i

Comp-Core/docs/latex/graph-kernel-paper-v2.tex

Read corpus pageOpen HTMLDownload HTML

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

Policy-Governed Context Slicing for Autonomous Agent Systems: A Lightweight Knowledge Graph Approach

Autonomous AI agent systems face a fundamental challenge: constructing reproducible, trustworthy context windows from large conversational histories while enforcing governance policies over what information may influence downstream decisions. We present the \textbf{Graph Kernel}, a deterministic context slicing engine implemented as a single Rust binary (${\sim}15$ KLOC) that combines a lightweight knowledge graph triple store with cryptographically-signed, policy-governed context window construction. Unlike genera

Comp-Core/docs/latex/graph-kernel-paper.tex

Read corpus pageOpen HTMLDownload HTML

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

Graph-Augmented Recursive Language Models for Personal Knowledge Systems

% ============================================================ We present \textbf{Cog-RLM}, a graph-augmented recursive language model architecture for personal knowledge systems that achieves 90.3\% accuracy on a comprehensive 103-question evaluation spanning ten cognitive dimensions, using a stock 3-billion parameter model with zero fine-tuning and zero inference cost. Our system extends the Recursive Language Model (RLM) paradigm~\citep{zhang2025rlm} with three novel contributions: (1)~a local knowledge graph pr

Comp-Core/packages/cognitive-twin/paper/latex/main.tex

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint structure candidatescore 100

Anticipation Geometry: Domain-General Trajectory Characterization with Knowledge Graph-Grounded Rewards

We present Anticipation Geometry, a mathematical framework that characterizes trajectories through arbitrary state spaces using seven geometric scalars: commitment, uncertainty, transition pressure, recovery margin, phase stiffness, novelty, and stability. Originally developed for physical motion capture in the Comp-Core system, we prove these scalars are domain-general, operating on any sequence of vectors in a metric space equipped with a differentiable time parameter. We combine this framework with knowledge gra

Comp-Core/papers/anticipation-geometry/paper.md

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

Computational Choreography: Deterministic Motion-to-Audio Synthesis via Geometric Anticipation Signals

We present Computational Choreography, a deterministic pipeline that transforms heterogeneous sensor input -- phone accelerometer, smartwatch heart rate, full-body IMU skeleton -- into real-time audio synthesis through geometric anticipation signals. The system guarantees deterministic replay: identical sensor input always produces identical audio output. The key innovation is the Anticipation Kernel, which computes seven geometric scalars (commitment, uncertainty, transition pressure, recovery margin, phase stiffn

Comp-Core/papers/computational-choreography/paper.md

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

Live Knowledge Graphs: Runtime Graph Integration for Continuous Domain Adaptation in Language Agents

Recent work on Domain-Specific Superintelligence (Belova et al., 2026) demonstrates that knowledge graph-derived training curricula produce domain specialists that outperform models 400x their size. However, this approach treats knowledge graphs as static training scaffolding: constructed once, used for fine-tuning, then discarded at inference. We present an alternative: runtime knowledge graph integration, where the graph is queried live during inference with provenance-tracked context slicing, real-time entity re

Comp-Core/papers/runtime-knowledge-graphs/paper.md

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

KARL: Advantage-Weighted Training from Full Agent Session Traces

Standard supervised fine-tuning (SFT) for language model agents operates on input-output pairs: a prompt and the response the model should produce. This format captures *what* an agent said but discards *why* it made specific decisions. We present KARL (Knowledge-Augmented Reinforcement Learning), a trajectory intelligence system that trains language model agents from full session traces rather than isolated completions. A trajectory in KARL records every tool call, file read, code edit, bash command, success signa

Comp-Core/papers/trajectory-intelligence/paper.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

Inscription-Conditioned Cognitive Twin: N'Ko Sigil Encoding as Semantic Compression for Long-Context Personality Models

Context window limitations constrain the fidelity of small personality models. A 4B parameter model with a 32K token context can hold roughly 8,000 words of conversation history before truncation begins discarding information critical to persona coherence. We present the Inscription-Conditioned Cognitive Twin (ICCT), an architecture that addresses this bottleneck by encoding conversation history as N'Ko inscriptions rather than English prose. The encoding uses 10 N'Ko sigils, each a single Unicode character derived

inscription-cognitive-twin-paper.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

KARL-Edge: Multi-Signal Reinforcement Learning for Software Engineering Agents on Commodity Hardware

We present KARL-Edge, an adaptation of the Knowledge Agents via Reinforcement Learning (KARL) framework to multi-tool software engineering agents running on commodity Apple Silicon hardware. Where the original KARL system (Chang et al., 2026) trains enterprise search agents using full off-policy RL with binary reward signals, our system introduces three architectural contributions: (1) a 5-signal composite reward function that decomposes trajectory quality into outcome, process, efficiency, verification, and consis

karl-research-paper.md

Read corpus pageOpen HTMLDownload HTML

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

Trajectory Memory Ledger

We present the Trajectory Memory Ledger, implemented in KARL, a schema-normalized experience replay system for improving AI coding agent performance through closed-loop feedback. The ledger records complete tool-use sequences during real coding sessions, normalizes them into an append-only schema, scores them using a six-signal composite reward function (outcome, process, efficiency, verification, consistency, and wasted motion), and uses the highest-scoring trajectories to generate advantage-weighted supervised fi

karl/paper/karl-paper.md

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint structure candidatescore 100

Live Knowledge Graphs: Runtime Graph Integration for Continuous Domain Adaptation in Language Agents

Recent work on Domain-Specific Superintelligence (Belova et al., 2026) demonstrates that knowledge graph-derived training curricula produce domain specialists that outperform models 400x their size. However, this approach treats knowledge graphs as static training scaffolding: constructed once, used for fine-tuning, then discarded at inference. We present an alternative: runtime knowledge graph integration, where the graph is queried live during inference with provenance-tracked context slicing, real-time entity re

live-knowledge-graphs/paper/paper.md

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint render candidatescore 100

Reading Tone from the Signal: Featural Acoustic Coding for Tone Resolution in N'Ko Speech Recognition

% Reading Tone from the Signal: % Featural Acoustic Coding for Tone Resolution in N'Ko Speech Recognition % Compiles with pdflatex (MacTeX). N'Ko shown via Unicode codepoints + % transliteration; IPA via tipa; architecture figure via TikZ.

nko-acoustic-coding/main.tex

Read corpus pageOpen HTMLDownload HTML

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

The Script That Machines Can't Read: Adapting Large Language Models for N'Ko

We present a systematic study of how large language models process N'Ko (\texttt{U+07C0--U+07FF}), an alphabetic script used by over 40 million Manding-language speakers in West Africa. Through activation profiling (``brain scanning'') of Qwen3-8B before and after fine-tuning, we demonstrate that: (1) fine-tuning concentrates N'Ko adaptation in the top 8 transformer layers, reducing activation magnitudes in reasoning layers while amplifying output confidence; (2) a three-stage training pipeline---continued pre-trai

nko-brain-scanner/arxiv-submission/main.tex

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

From Dead Circuits to Living Speech: Activation Profiling and Script-Native ASR for N'Ko

N'Ko is an alphabetic script serving over 40 million Manding-language speakers across West Africa, engineered by Solomana Kant\'{e} in 1949 with a strict 1:1 phoneme-to-character mapping, explicit tonal diacritics, and zero spelling exceptions. We present a dual-thread investigation into why large language models (LLMs) fail on N'Ko and how to build audio-to-N'Ko speech recognition that bypasses LLMs entirely. \textbf{Thread 1 (Diagnostic):} We perform activation profiling---a ``brain scan''---of Qwen2-72B-Instruct

nko-brain-scanner/paper/archive/main_v2.tex

Read corpus pageOpen HTMLDownload HTML

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

The Script That Machines Can't Read: Adapting Large Language Models for N'Ko

We present a systematic study of how large language models process N'Ko (\texttt{U+07C0--U+07FF}), an alphabetic script used by over 40 million Manding-language speakers in West Africa. Through activation profiling (``brain scanning'') of Qwen3-8B before and after fine-tuning, we demonstrate that: (1) fine-tuning concentrates N'Ko adaptation in the top 8 transformer layers, reducing activation magnitudes in reasoning layers while amplifying output confidence; (2) a three-stage training pipeline---continued pre-trai

nko-brain-scanner/paper/archive/main.tex

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

From Dead Circuits to Living Speech: Activation Profiling, Script-Native Architecture Search, and Finite-State Phonotactics for N'Ko Automatic Speech Recognition

\nko{} is an alphabetic script serving over forty million Manding-language speakers across West Africa, engineered by Solomana Kant\'e in 1949 with a strict one-to-one phoneme-to-character mapping, explicit tonal diacritics, and zero spelling exceptions. We present a dual-thread investigation into why large language models fail on \nko{} and how to construct audio-to-\nko{} speech recognition that bypasses such models entirely. In the diagnostic thread, we perform activation profiling of Qwen2-72B-Instruct (4-bit N

nko-brain-scanner/paper/archive/nko_paper_v3.tex

Read corpus pageOpen HTMLDownload HTML

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

Theorems, Proofs, and Derivations for N'Ko Script-Native ASR

This document collects the formal mathematical results underlying the N'Ko Brain Scanner and ASR system. We present five main theorems with proofs, three derivations of key quantities, and two corollaries that connect the LLM diagnostic thread to the ASR construction thread. The results establish: (1) a phonetic transparency advantage for CTC decoding on bijective scripts, (2) bounds on the translation tax in under-represented scripts, (3) completeness and soundness of the FSM phonotactic validator, (4) a circuit d

nko-brain-scanner/paper/archive/nko_theorems.tex

Read corpus pageOpen HTMLDownload HTML

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint structure candidatescore 100

N'Ko as an Extensible Phonemic Substrate for Governed Low-Resource Speech

Low-resource speech systems usually fail twice: first because there is not enough audio/text data, and second because the available evaluation scripts do not preserve the phonemic structure of the language being measured. This paper argues that N'Ko offers a different path. Because N'Ko is a phonetic, right-to-left script designed for Manding languages and equipped with tone, nasalization, and documented foreign-sound diacritics, it can function as an extensible phonemic substrate: a deterministic sound-code for co

nko-brain-scanner/paper/current/nko_phonemic_substrate_paper.md

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Convert into the standard paper schema, add citations, and render a draft PDF.

working paperpreprint render candidatescore 100

N'Ko as Computational Infrastructure: Script-Native Speech Recognition, a Phonemically Interpretable Error Metric, and Admissible Tone Correction

\usepackage[margin=1in]{geometry} \usepackage{booktabs} \usepackage{array} \usepackage{amsmath} \usepackage{amssymb} \usepackage{graphicx} \usepackage{hyperref} \usepackage[numbers]{natbib} \usepackage{xcolor} \usepackage{longtable} \usepackage{caption} \usepackage{seqsplit}

nko-brain-scanner/paper/current/paper_canonical_nko_agp_20cer.tex

Read corpus pageOpen HTMLDownload HTML

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

Dead Circuits: Activation Profiling and Script Invisibility in Large Language Models

Large language models achieve remarkable performance on languages written in Latin, Cyrillic, CJK, and Arabic scripts. We ask what happens when these models encounter a script that is absent from their pre-training data. We perform activation profiling---a per-layer ``brain scan''---of Qwen3-8B processing 100 parallel English/N'Ko sentence pairs. N'Ko is an alphabetic script serving over 40 million Manding-language speakers across West Africa, engineered in 1949 with a strict phoneme-to-grapheme bijection, explicit

nko-brain-scanner/paper/current/paper1_dead_circuits.tex

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

Living Speech: Script-Native Automatic Speech Recognition for N'Ko

\documentclass[11pt]{article} \usepackage{acl} \usepackage{times} \usepackage{latexsym} \usepackage{graphicx} \usepackage{booktabs} \usepackage{amsmath} \usepackage{amssymb} \usepackage{hyperref} \usepackage{multirow} \usepackage{xcolor} \usepackage{enumitem} \usepackage{tipa}

nko-brain-scanner/paper/current/paper2_living_speech.tex

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

Script Invisibility Is Structural: Activation Profiling Across Three LLM Families

A prior study demonstrated that Qwen3-8B processes N'Ko text with severely diminished neural activation compared to English, a phenomenon termed \emph{script invisibility}. That finding left an open question: is the deficit specific to one model, or is it a structural property of all models trained on corpora where N'Ko is absent? We answer this by performing identical activation profiling---per-layer extraction of L2 norm, Shannon entropy, sparsity, and kurtosis---on three architecturally distinct models: Qwen3-8B

nko-brain-scanner/paper/current/paper3_cross_model.tex

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

working paperpreprint render candidatescore 100

Beyond Controlled Comparison: Deployment Properties of Script-Aware ASR for N'Ko

Controlled experiments show that phonetically transparent scripts yield lower CER for CTC-based ASR. But ASR systems are not evaluated in controlled conditions---they encounter unseen vocabulary, new speakers, and domain shift. This paper assembles deployment-relevant evidence for Bambara ASR systems using N'Ko (bijective script) and Latin (many-to-many script), anchored by the verified 20.57\% N'Ko trajectory checkpoint but drawing on both current and historical experiments. First, \textbf{compositional generaliza

nko-brain-scanner/paper/current/paper5_deployment.tex

Read corpus pageOpen HTMLDownload HTMLOpen curated reader

Next: Compile/render the source, verify references and figures, then add to the curated atlas.

Technical lane

Architectures and systems

These should not all become academic papers, but they are the implementation substrate: system maps, runbooks, invariants, and architecture documents that prove where a claim came from.

Evidence lane

Experiment writeups

These are where proof should attach: run IDs, metrics, datasets, screenshots, logs, notebooks, replay commands, and failure modes.

Full intake

Everything grouped by program

This is intentionally broad. The purpose is to stop losing work: each entry carries a source anchor, readiness label, and next action so future agents can promote or reject it explicitly.

ProgramAgents That Account for Themselves532 entries
ProgramEmbodied Trajectory Systems317 entries
ProgramLanguage as Infrastructure238 entries
ProgramBusiness Systems20 entries
ProgramResearch Backlog28 entries
ProgramResearch Practice5 entries
ProgramProtocol and Compute7 entries