Typed states for the displayed outbound observations.
Source: paper_references, paper_reference_links, observed 2026-08-06T23:26:57.754155Z
Paper Citation Record · LEDGER
As of 7 August 2026, this Paper Citation Record lists 97 of 97 outbound references and 1 inbound Pith citation observation for arXiv:2506.17846.
A citation records a reference. It does not transfer a finding from one paper to another.
Typed states for the displayed outbound observations.
Source: paper_references, paper_reference_links, observed 2026-08-06T23:26:57.754155Z
One-hop event checks from named stored sources.
Source: scholarly_work_events, retraction_status_cache, observed 2026-08-07T06:34:17.273281+00:00
Pith citing papers itemized under the disclosed page cap.
Source: paper_references, paper_reference_links, observed 2026-08-06T19:27:34.770229Z
A source-named dated measurement, never combined with another source.
Source: pith, observed 2026-08-06T19:27:35.026558Z
97 of 97 outbound references displayed
External citation measurements
No source-named external measurement is stored.
Observation ab12fc43-486f-4daa-bac5-8049aee7dcf1 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) AI Control: Improving Safety Despite Intentional Subversion
Reference 1
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 0075e0b6-928c-4ab7-81c9-2173cc5bde02 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) A sketch of an AI control safety case
Reference 2
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 3fdd020e-e755-40fa-90bc-d03e4a508367 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Subversion Strategy Eval: Can language models statelessly strategize to subvert control protocols?
Reference 3
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 7c3799c3-51b8-44de-a3a1-0bf9ccc84549 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Adaptive Deployment of Untrusted LLMs Reduces Distributed Threats
Reference 4
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6bdb94a0-4680-491c-9e79-36571991cb74 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Hidden in plain text: Emergence & mitigation of steganographic collusion in LLMs
Reference 5
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 09f8f5fc-d200-4615-8aee-44edd28c4f68 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Mechanistic anomaly detection and elk
Reference 6
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 7a0f4489-f5b4-470c-9219-226c2b6fdc6f · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Representation Engineering: A Top-Down Approach to AI Transparency
Reference 7
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 40e474c2-869f-4a60-8bac-dba478f0581b · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Open Problems in Mechanistic Interpretability
Reference 8
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 20397426-c115-47a5-a529-d2f1b29d540d · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Berkovitz and Negash G
Reference 9
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 30a29f39-a2b9-47a2-864d-13cd9eca1f87 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) A Research Agenda: Dynamic Models to Defend Against Correlated Attacks
Reference 10
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 8f4e846f-ed77-416d-bbeb-9865ddf0fd5a · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Learning to summarize from human feedback
Reference 11
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation b24e0c39-92c9-420b-9a4a-bd334de653e3 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Training language models to follow instructions with human feedback
Reference 12
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 9dc2a1c7-2f43-4554-b9dd-3faf7f7c8e44 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Fine-Tuning Language Models from Human Preferences
Reference 13
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 28b69cd8-196e-450e-a172-3ea8edad8196 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) A General Language Assistant as a Laboratory for Alignment
Reference 14
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6e751683-b89e-488a-80fe-3877b08c11a4 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Constitutional AI: Harmlessness from AI Feedback
Reference 15
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 5e7e2798-2cdb-42ae-a2e2-748c3cded518 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Cooperative inverse reinforcement learning
Reference 16
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation a75905cd-e4ab-40db-969b-80a1713f02c9 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Scalable agent alignment via reward modeling: a research direction
Reference 17
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 10908b4d-a984-4b7c-980f-9e872c76bd49 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned
Reference 18
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation c5dfe7dc-112a-468e-9f77-317055abfb9a · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) The Pristine survey -- XXII. A serendipitous discovery of an extremely Li-rich very metal-poor giant and a new method of $^6$Li/$^7$Li isotope measurement
Reference 19
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 6ec8cef2-c81f-4cae-87c5-0b2407571081 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Consequences of misaligned AI
Reference 20
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 5775113a-50f0-4329-8b43-f4bba1509c50 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Risks from Learned Optimization in Advanced Machine Learning Systems
Reference 21
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation b7279a7c-6550-4fcf-8609-dc0f7a0a6616 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) On the relevance of the Godot Engine in the indie game development industry
Reference 22
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 523ab9c4-8d1d-4b4b-b703-f3731febc4bf · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Steganography in chain of thought reasoning
Reference 23
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation ec1ebec8-65a9-4a13-bcb2-5a02e34890b0 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Neural linguistic steganography
Reference 24
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation f152e7d6-8fe4-401f-afad-b3018f80b3bb · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Information hiding-a survey
Reference 25
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 1afc3dff-0957-4efc-945d-c63465a950a6 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) CycleGAN, a Master of Steganography
Reference 26
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6937da5d-d4df-4bee-bbda-49f5a95c1e45 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) A review on text steganography techniques
Reference 27
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation d1391414-55f8-42d2-89f3-84236a15ee36 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Emergent misalignment: Narrow finetuning can produce broadly misaligned llms
Reference 28
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation be5e32ca-d5bb-40de-99c1-b3047f8595bc · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Discovering latent knowledge in language models without supervision
Reference 29
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 339b3efe-c9c7-4a3c-9997-1077bf4cd380 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Arc’s first technical report: Eliciting latent knowledge
Reference 30
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation fe61fc05-1d5e-429f-bd1e-50696299657a · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Is Power-Seeking AI an Existential Risk?
Reference 31
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 0c43d34d-c830-4c9c-95f8-cd70bdf42d44 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Optimal policies tend to seek power
Reference 32
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 91d1d0cd-e82a-455e-9c32-91d9b55043f2 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Corrigibility
Reference 33
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation c4d5ef54-f1ca-4500-93fa-7c05a3fce28c · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) AI safety via debate
Reference 35
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 0e8cf767-1101-4c6e-92c4-bc2ccc28eeaf · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Debate Helps Supervise Unreliable Experts
Reference 36
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 4cb71390-ee26-42d7-b8bd-9ba263b7c9da · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Towards Guaranteed Safe AI: A Framework for Ensuring Robust and Reliable AI Systems
Reference 37
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6484fdc9-805a-4106-bfc3-6eced7d4b468 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) A new robust approach for the polytomous logistic regression model based on R\'enyi's pseudodistances
Reference 38
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 1de1f1e8-f83b-48e8-b322-b84b1cdf1473 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) The Effects of Reward Misspecification: Mapping and Mitigating Misaligned Models
Reference 39
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 972a95bd-1328-4030-a171-307a97df5f0c · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) The Alignment Problem from a Deep Learning Perspective
Reference 40
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 59348b0c-9cde-4e9e-9ea7-da61df4f85c2 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Deep reinforcement learning from human preferences
Reference 41
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 89095439-ec12-4800-a7a6-dca73a60ef5d · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) One-Year In: COVID-19 Research at the International Level in CORD-19 Data
Reference 42
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 329a6965-01d5-4c3c-85e1-d48057160d83 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Measuring Progress on Scalable Oversight for Large Language Models
Reference 43
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 07886c10-0fc8-4fbb-a355-0f241ea5eda3 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Discovering Language Model Behaviors with Model-Written Evaluations
Reference 44
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation b4d26c89-ef97-4f54-a784-e949f76de09d · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Steering Llama 2 via Contrastive Activation Addition
Reference 45
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation e963b313-e3ce-4a8f-a981-2ad6bf27bafc · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Truthful AI: Developing and governing AI that does not lie
Reference 46
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 75d5ecf9-2be6-49fc-90d7-fe55ee7d10b2 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Safe reinforcement learning via shielding
Reference 47
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 118426b5-8c24-4398-b2a4-93a8eb3412ae · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Robust adversarial reinforcement learning
Reference 48
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 4f938f2e-83ea-4ff4-837f-3ea38260fa2d · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Dynamic programming and optimal control, volume 1
Reference 49
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 0a2f84d3-4ddc-4f66-81e1-a5834221f5dd · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Robust and optimal control
Reference 50
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation ede4711e-9d33-4c3e-bdd8-d0690d0accaf · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Optimal control theory: an introduction
Reference 51
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 137e513d-2f59-42df-8ad6-7f2ee19dddd1 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Optimal control
Reference 52
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation b543f7f1-88cf-4637-9fff-b266e9ee4bdf · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Adaptive control
Reference 53
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 0c6359aa-fbd5-4a43-b3ae-6268e04a0564 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Model predictive control
Reference 54
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 48cd507c-b336-4db8-be0b-8ff796034a38 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Stochastic optimal control: the discrete-time case , volume 5
Reference 55
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation d6702858-719b-42ad-8e71-e6cbd3901a41 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Nonlinear control systems
Reference 56
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation d0662494-ffad-4fc4-824f-a158065e2adb · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Geometric control of mechanical systems: modeling, analysis, and design for simple mechanical control systems
Reference 57
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation eb062195-036f-4e8d-9510-bd13859555ac · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Hybrid dynamical systems: modeling, stability, and robustness
Reference 58
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 1afd2908-f842-4578-8f7d-eae8cea6b095 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Reinforcement learning: An introduction
Reference 59
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 85910cfd-120d-43a4-bf1c-259ce9df6602 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) ˙x∗(t) = ∂H ∂λ (t, x∗(t), u∗(t), λ(t)) (12) ˙λ(t) = − ∂H ∂x (t, x∗(t), u∗(t), λ(t)) (13) where the Hamiltonian H(t, x, u, λ) = λ0f 0(t, x, u) + Pn i=1 λi(t)f i(t, x, u)
Reference 60
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 6aec0a9b-d173-4222-8f53-86f322499486 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) For almost all t ∈ [t0, t1], H(t, x∗(t), u∗(t), λ(t)) = sup z∈Ω(t,x∗(t)) H(t, x∗(t), z, λ(t)) (14)
Reference 61
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 21d58aa7-2b20-42cd-9efd-47d8fd24fa3f · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) At the endpoints (t0, x∗(t0)) and (t1, x∗(t1)), the vector (−H(t0), λ(t0), H(t1), −λ(t1)) (where H(t) is H evaluated along the extremal) is orthog- onal to the end-manifold B
Reference 62
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 81438081-e24b-4463-9059-09359952aa4f · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Unresolved cited work
Reference 63
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 795b0bda-e112-476c-bc4c-209e73382294 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Unresolved cited work
Reference 64
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 01562de8-af0b-45bc-82d5-17daae9e2a60 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) This involves adjusting operational parameters of physical components
Reference 65
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation b39cf135-32c6-441b-ab87-3afd32da7477 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) This concerns distributing computational tasks across available processing units (e.g., GPU kernels)
Reference 66
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation b5ca32a1-ea19-4f32-99f2-8ff8c5a68446 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) This includes mechanisms like Error Correction Codes (ECC) in memory or redundant array of inexpensive disks (RAID) for storage
Reference 67
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 5260a0b9-0003-4c11-8515-beabdeac6f29 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) OS and containerization tools allow setting limits on resource consumption per process or container (e.g., Linux cgroups)
Reference 68
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 25bcc983-12b4-44ec-bbe9-e627e225b119 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) The OS scheduler manages the execution order of processes and threads
Reference 69
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 568f8085-15ae-4e6b-bd49-e1d2c977be39 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Using containers, virtual machines, or specific OS mechanisms (e.g., seccomp, AppArmor) to restrict the privileges and visibility of AI processes
Reference 70
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation c41fae4d-9ee8-4b4c-8e89-5958c1c27db3 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Frameworks often perform optimisations like op- erator fusion, constant folding, and memory planning on the AI model’s computational graph
Reference 71
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 79952eb8-3a97-4d05-8ba2-aed4f906db05 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Techniques applied during computation to maintain numeri- cal precision and prevent issues like vanishing/exploding gradients
Reference 72
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 03a971ba-78f7-47db-8445-760c7c2d4015 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Mechanisms for managing data and model parallelism, synchronizing gradients, and handling parameter updates across multiple devices or nodes
Reference 73
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 2a14673b-1ee7-446f-87da-5b782d41d6b8 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Its efficiency and correctness are crucial for successful model development
Reference 74
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 0207fbd2-b062-4de4-a1d9-ddc2c9a2b47f · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) The deliberate selection of network structures, layer types, and connectivity
Reference 75
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 6f494e16-29ea-4065-a968-b5280365186d · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Incorporating architectural elements that intrinsi- cally promote better generalisation or robustness, such as dropout layers or batch normalisa- tion
Reference 76
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation bd8be8eb-0581-47a3-83f5-d6710a09d4d5 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Methods like pruning (removing less important weights/neurons) and quantization (reducing the precision of weights and activations)
Reference 77
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 71f2ee9b-fd11-49bb-a2c3-054be200ae91 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Selecting, cleaning, and transforming the training data
Reference 78
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 24b7d4b6-37a5-481b-822e-533363ff6250 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Defining the loss function Ltask that the model aims to minimise, and potentially adding regularization terms Lregi or constraints
Reference 79
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 51470257-ea07-4e3a-9e00-eee556489701 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Choosing the optimizer (e.g., SGD, Adam) and its parameters (e.g., learning rate ηt, batch size Bt, momentum βt)
Reference 80
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 683017bc-a545-444d-9d06-72ce2f30305e · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Monitoring performance on a validation set and stop- ping training when performance no longer improves, saving model checkpoints periodically
Reference 81
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation de7b95c0-a8b2-4dd9-9fe6-a012e10dd32d · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Post-processing model outputs to detect and remove or modify undesirable content (e.g., harmful language, private information)
Reference 82
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 61463965-6dd4-4ebf-a701-7cd32e43c7db · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Abstaining from providing an output or escalating to a human supervisor if the model’s confidence in its output is below a certain threshold
Reference 83
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation d636dba1-dc82-4894-b1d9-e124cb2b2de6 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Unresolved cited work
Reference 84
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation e9b295af-af6a-4be9-9f08-84368c33ec4c · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Altering or ablating spe- cific internal parts of a model (e.g., neurons, attention heads, identified circuits) that are understood to contribute to certain behaviours
Reference 85
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 33e4d8b3-b717-4388-a6c0-b7e52fd0e5d1 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Using explanations of model behaviour (e.g., why a particular output was generated) to identify and diagnose issues
Reference 86
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation be31639d-5d89-4ecb-abee-769f7f19d1b0 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Directly modifying model weights or structure based on interpretability insights to correct specific knowledge or behaviours without full retraining
Reference 87
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 3771ab7b-4652-4736-85e8-0a32f8be8af0 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Techniques for capturing human preferences and values to guide AI behaviour, often by training a reward model RM (y) based on human feedback
Reference 88
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 5f6db323-c4bd-4170-bb47-0db365045fbd · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Using a learned reward model or direct human feedback to fine-tune a base AI model towards desired behaviours
Reference 89
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 0528181c-23ff-4844-8914-5c5de2637eb5 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Systematically probing AI systems to discover vulnerabilities, failure modes, and misalignments before deployment
Reference 90
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation be72f600-06fa-4be9-a576-4eee69244c81 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Defining explicit principles, rules, or constitutions that the AI must adhere to, and using AI itself to critique and revise its outputs to comply with these rules
Reference 91
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation f6f45057-65f6-4903-86da-4d7e61b71c53 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Designing the rules of interaction, com- munication protocols, and reward/utility functions for individual agents within a multi-agent system
Reference 92
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation a21a1b03-6151-4c23-8997-e89952978811 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Implementing systems for establish- ing, communicating, monitoring, and enforcing social norms or rules within an AI agent population
Reference 93
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation c8a6f8aa-4642-4ae1-b1a5-83fbe2dacebf · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Establishing standardized ways for agents to exchange information, signal intent, and coordinate actions
Reference 94
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 0f60556e-98cf-449a-ad22-b3078b03615a · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Governments and international bodies establishing laws, standards, and policies governing the development, deployment, and use of AI
Reference 95
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation c3ed02ca-6c23-420c-ab9c-fb9993b972d3 · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Professional organiza- tions, research institutions, and industry consortia developing and promoting principles for responsible AI
Reference 96
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 1f3a2255-c9e2-4c71-af75-151daf40becd · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Efforts to inform the public about AI, foster critical discussion, and involve diverse stakeholders in shaping AI’s future
Reference 97
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation 6e1b6198-ad87-4f12-860d-c8be1e6c09ab · outbound
Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) Formal processes for evaluating the potential societal impacts of AI systems before and during deployment, and for independently verifying claims about their safety or fairness
Reference 98
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.
Observation ae3eb31d-6151-4fda-9b35-fb702528aa8a · inbound
Towards Measurement Theory for Artificial Intelligence Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack)
Reference 33
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.