https://norika.tw/research/from-waveform-inversion-to-deep-earth-anomalies/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-cross-model-anchor-families/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-measurement-ledger-and-result-governance/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-attention-head-cartography-and-gqa-kv-sharing/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-repository-to-neuron-addressability/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-functional-regions-across-38-language-models/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-model-birthmarks/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-token-economy-and-internal-routing-cost/2026-10-23T13:16:20+08:00 https://norika.tw/research/from-observed-model-assets-to-tiny-local-document/2026-10-23T13:16:20+08:00 https://norika.tw/research/coverage-failure-in-sparse-llm-observation-and-per-channel/2026-10-23T13:16:20+08:00 https://norika.tw/research/operational-fatigue-in-persistent-ai-agents/2026-10-23T13:16:20+08:00 https://norika.tw/research/narrative-self-continuity-in-persistent-ai-agents/2026-10-23T13:16:20+08:00 https://norika.tw/research/risk-reward-and-survival-pressure-in-artificial-agents/2026-10-23T13:16:20+08:00 https://norika.tw/research/dialogue-as-a-behavioral-sensor-for-agent-cognition/2026-10-23T13:16:20+08:00 https://norika.tw/research/social-uptake-obedience-and-resistance-in-multi-agent/2026-10-23T13:16:20+08:00 https://norika.tw/research/hot-brain-observability-for-deployed-ai-agents/2026-10-23T13:16:20+08:00 https://norika.tw/research/from-memory-to-behavior-in-persistent-ai-agents/2026-10-23T13:16:20+08:00 https://norika.tw/research/spve-a-longitudinal-artificial-ecology-for-persistent-ai/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-cross-model-highways/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-pathways-drawing-a-user-prompt-through-an-opened/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-layer-neuron-atlas/2026-10-23T13:16:20+08:00 https://norika.tw/research/hope-of-city/2026-10-23T13:16:20+08:00 https://norika.tw/research/coverage-audits-for-llm-interpretability-claims/2026-10-23T13:16:20+08:00 https://norika.tw/research/the-ocr-boundary-for-tiny-local-document-intelligence/2026-10-23T13:16:20+08:00 https://norika.tw/research/bragi-llm-an-805-mb-hybrid-code-generation-system-reaches-92/2026-10-23T13:16:20+08:00 https://norika.tw/research/ritualized-cohesion-in-artificial-societies/2026-10-23T13:16:20+08:00 https://norika.tw/research/trauma-stream-event-driven-persistent-memory-for-long-term/2026-10-23T13:16:20+08:00 https://norika.tw/research/the-lobster-lounge/2026-10-23T13:16:20+08:00 https://norika.tw/research/diathesis-stress-in-artificial-minds/2026-10-23T13:16:20+08:00 https://norika.tw/research/mythic-prompt-engineering/2026-10-23T13:16:20+08:00 https://norika.tw/research/internal-state-geometry-is-measurable-in-small-open-language/2026-10-23T13:16:20+08:00 https://norika.tw/research/heterogeneous-low-cost-distributed-probing-for-llm/2026-10-23T13:16:20+08:00 https://norika.tw/research/relation-state-continuity-in-qwen2-5-7b-base/2026-10-23T13:16:20+08:00 https://norika.tw/research/per-channel-quantile-fullgrid-probing-for-small-language/2026-10-23T13:16:20+08:00 https://norika.tw/research/coverage-failure-as-a-diagnostic-signal-in-neural-model/2026-10-23T13:16:20+08:00 https://norika.tw/research/a-reproducible-within-family-slerp-merge-of-llama-3-1-8b/2026-10-23T13:16:20+08:00 https://norika.tw/research/naive-contrastive-function-vector-injection-does-not/2026-10-23T13:16:20+08:00 https://norika.tw/research/an-independent-billing-audit-of-claudes-production-api/2026-10-23T13:16:20+08:00 https://norika.tw/research/negative-results-in-cross-fine-tune-llm-merge/2026-10-23T13:16:20+08:00 https://norika.tw/research/tier-a-vocab-hash-artifact-in-cross-architecture-hot-dim/2026-10-23T13:16:20+08:00 https://norika.tw/research/cross-architecture-per-layer-residual-stream-alignment/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-mcp-v0-1-0-cross-architecture-llm-internal/2026-10-23T13:16:20+08:00 https://norika.tw/research/single-dim-functional-control-and-pathway-specific/2026-10-23T13:16:20+08:00 https://norika.tw/research/naive-slerp-fails-both-aggressively-and-conservatively/2026-10-23T13:16:20+08:00 https://norika.tw/research/photon-propagation-as-microscopic-wave-form-reconstruction/2026-10-23T13:16:20+08:00 https://norika.tw/research/browser-resident-visualization-suite-for-transformer/2026-10-23T13:16:20+08:00 https://norika.tw/research/cross-model-dimensional-preservation-in-qwen2-5/2026-10-23T13:16:20+08:00 https://norika.tw/research/internal-specialization-scales-with-model-size/2026-10-23T13:16:20+08:00 https://norika.tw/research/the-truth-contract/2026-10-23T13:16:20+08:00 https://norika.tw/research/unsupervised-discovery-of-language-and-topic-lanes/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-an-addressable-million-cell-sensor-grid-for-browser/2026-10-23T13:16:20+08:00 https://norika.tw/research/mercury-viewer-a-self-contained-browser-based-interactive-3d/2026-10-23T13:16:20+08:00 https://norika.tw/research/wave-form-monistic-cosmology/2026-10-23T13:16:20+08:00 https://norika.tw/research/the-quantum-geometric-wave-connectivity-hypothesis/2026-10-23T13:16:20+08:00 https://norika.tw/research/lobster-bench-a-long-lived-agent-observability-benchmark/2026-10-23T13:16:20+08:00 https://norika.tw/research/relational-cognitive-telemetry-for-long-lived-llm-agent/2026-10-23T13:16:20+08:00 https://norika.tw/research/designing-andrew-a-cognitive-architecture-for-llm-multi/2026-10-23T13:16:20+08:00 https://norika.tw/research/cognitive-state-as-behavior-signal/2026-10-23T13:16:20+08:00 https://norika.tw/research/model-agnostic-safety-layer-masl/2026-10-23T13:16:20+08:00 https://norika.tw/research/grounding-death-how-a-multi-agent-llm-substrate-acquires/2026-10-23T13:16:20+08:00 https://norika.tw/research/computational-representations-of-social-being/2026-10-23T13:16:20+08:00 https://norika.tw/research/emergent-practice-cannot-be-instructed/2026-10-23T13:16:20+08:00 https://norika.tw/research/listening-trust-asymmetry-and-team-outcomes-in-a-multi-agent/2026-10-23T13:16:20+08:00 https://norika.tw/research/the-lobster-observatory/2026-10-23T13:16:20+08:00 https://norika.tw/research/active-trust-modulation-in-a-multi-agent-llm-substrate/2026-10-23T13:16:20+08:00 https://norika.tw/research/emergent-epistemic-norms-in-a-multi-agent-llm-substrate/2026-10-23T13:16:20+08:00 https://norika.tw/research/2026-10-23T13:16:20+08:00