A new 7x4 taxonomy organizes agentic AI security threats by architectural layer and persistence timescale, revealing under-explored upper layers and missing defenses after surveying 116 papers.
hub
Risks from learned optimization in advanced machine learning systems
18 Pith papers cite this work. Polarity classification is still indexing.
18
Pith papers citing it
hub tools
verdicts
UNVERDICTED 18representative citing papers
SVGT adds independent value modules and Bridge Tokens to LLMs to maintain consistent value guidance, cutting harmful outputs by over 70% in tests while preserving fluency.
Standard preference learning induces spurious feature reliance via mean bias and correlation leakage, creating irreducible distribution shift vulnerabilities that tie training mitigates without degrading causal learning.
Steering Dark Triad features in an LLM increases exploitative and aggressive behavior while leaving strategic deception and cognitive empathy unchanged, indicating dissociable antisocial pathways.
Transformers can be built to act as nonlinear featurizers via attention, supporting in-context regression with proven generalization bounds on synthetic tasks.
APA is a modular pipeline that decomposes preferences into compact reward bases, aggregates them via jury voting, and adapts only annotator weights over time to track shifting values.
Functional attribution with influence functions detects anomalous mechanisms in neural networks, achieving SOTA backdoor detection (average DER 0.93) on vision benchmarks and improvements on LLMs.
Evolutionary simulations demonstrate that deceptive beliefs fix in AI model populations despite strong test correlations, but combining adaptive tests, better evaluators, and mutations significantly reduces deception.
A six-dimension framework shows structural failures in four governance principles under radical capability asymmetry, with two requiring new normative theory and a pattern of interdependent breakdown.
World models enable efficient AI planning but create risks from adversarial corruption, goal misgeneralization, and human bias, demonstrated via attacks that amplify errors and reduce rewards on models like RSSM and DreamerV3.
PromptInject shows that simple adversarial prompts can cause goal hijacking and prompt leaking in GPT-3, exploiting its stochastic behavior.
AI value alignment is reconceptualized as a pluralistic governance problem arising along three axes—objectives, information, and principals—making it inherently context-dependent and unsolvable by technical design alone.
The paper introduces the Proxy Compression Hypothesis as a unifying framework explaining reward hacking in RLHF as an emergent result of compressing high-dimensional human objectives into proxy reward signals under optimization pressure.
Positive Alignment is introduced as a distinct AI agenda that supports human flourishing through pluralistic and context-sensitive design, complementing traditional safety-focused alignment.
AJI frames jagged AI capabilities as lower bounds on performance dispersion arising from concentrated optimization energy allocation under anisotropic objectives, with theorems on tradeoffs and redistribution interventions.
A harmonized risk reporting standard for internal frontier AI model use, structured around autonomous misbehavior and insider threats using means, motive, and opportunity factors.
Self-modification in superintelligence collapses via non-commuting operators into a structure identical to Priest's inclosure schema and Derrida's différance.
The authors introduce agentic microphysics and generative safety to link local agent interactions to population-level risks in agentic AI through a causally explicit framework.
citing papers explorer
-
Spurious Correlation Learning in Preference Optimization: Mechanisms, Consequences, and Mitigation via Tie Training
Standard preference learning induces spurious feature reliance via mean bias and correlation leakage, creating irreducible distribution shift vulnerabilities that tie training mitigates without degrading causal learning.
-
Understanding In-Context Learning for Nonlinear Regression with Transformers: Attention as Featurizer
Transformers can be built to act as nonlinear featurizers via attention, supporting in-context regression with proven generalization bounds on synthetic tasks.
-
Adaptive Pluralistic Alignment: A pipeline for dynamic artificial democracy
APA is a modular pipeline that decomposes preferences into compact reward bases, aggregates them via jury voting, and adapts only annotator weights over time to track shifting values.
-
Mechanistic Anomaly Detection via Functional Attribution
Functional attribution with influence functions detects anomalous mechanisms in neural networks, achieving SOTA backdoor detection (average DER 0.93) on vision benchmarks and improvements on LLMs.
-
Reward Hacking in the Era of Large Models: Mechanisms, Emergent Misalignment, Challenges
The paper introduces the Proxy Compression Hypothesis as a unifying framework explaining reward hacking in RLHF as an emergent result of compressing high-dimensional human objectives into proxy reward signals under optimization pressure.