arXiv:2608.24662v1 Announce Type: new
Abstract: Large language models (LLMs) are commonly evaluated under the assumption that their observable behavior is primarily determined by model weights, training data, alignment procedures, and user prompts. This view is incomplete. Modern inference pipelines may systematically modify the probability distribution produced by a model immediately before token selection, creating an additional layer of control between frozen weights and observed text.
While controlled generation (e.g., PPLM, GeDi, DExperts, FUDGE) and text-watermarking systems (e.g., SynthID-Text) demonstrate the technical maturity of decoding- and logit-level interventions, the governance, security, and economic implications of an undisclosed inference policy remain comparatively underexplored. This paper examines the emergence of inference-time framing bias: the systematic modification of generated language toward political, ideological, institutional, or commercial frames via interventions applied after model inference but before token sampling.
We formalize the operational reality Model != Deployed System and introduce three concepts: (1) the Inference Attribution Problem, characterizing why observed behavioral bias cannot generally be causally attributed to model weights alone under limited observability; (2) Probability Placement, defining a hypothetical advertising primitive in which commercial influence is implemented through systematic shifts in generation probabilities rather than explicit product insertions; and (3) Inference Policy Transparency, a governance principle for making deployment-layer interventions auditable. We examine these concepts in relation to Article 5 of the EU AI Act, the EU Digital Services Act, and FTC doctrines.
Abstract: Large language models (LLMs) are commonly evaluated under the assumption that their observable behavior is primarily determined by model weights, training data, alignment procedures, and user prompts. This view is incomplete. Modern inference pipelines may systematically modify the probability distribution produced by a model immediately before token selection, creating an additional layer of control between frozen weights and observed text.
While controlled generation (e.g., PPLM, GeDi, DExperts, FUDGE) and text-watermarking systems (e.g., SynthID-Text) demonstrate the technical maturity of decoding- and logit-level interventions, the governance, security, and economic implications of an undisclosed inference policy remain comparatively underexplored. This paper examines the emergence of inference-time framing bias: the systematic modification of generated language toward political, ideological, institutional, or commercial frames via interventions applied after model inference but before token sampling.
We formalize the operational reality Model != Deployed System and introduce three concepts: (1) the Inference Attribution Problem, characterizing why observed behavioral bias cannot generally be causally attributed to model weights alone under limited observability; (2) Probability Placement, defining a hypothetical advertising primitive in which commercial influence is implemented through systematic shifts in generation probabilities rather than explicit product insertions; and (3) Inference Policy Transparency, a governance principle for making deployment-layer interventions auditable. We examine these concepts in relation to Article 5 of the EU AI Act, the EU Digital Services Act, and FTC doctrines.
This article has been indexed from cs.AI updates on arXiv.org
Read the original article: