Inference with Constrained Hidden Markov Models in PRISM

Viterbi training in PRISM

VT (Viterbi training), or hard EM, is an efficient way of parameter learning for probabilistic models with hidden variables. Given an observation $y$, it searches for a state of hidden variables $x$ that maximizes $p(x,y \mid θ)$ by coordinate ascent on parameters $θ$ and $x$. In this paper we introduce VT to PRISM, a logic-based probabilistic modeling system for generative models. VT improves PRISM in three ways. First VT in PRISM converges faster than EM in PRISM due to the VT's termination condition. Second, parameters learned by VT often show good prediction performance compared to those l…

PRISM：从语言模型激活中恢复指令集

PRISM: Recovering Instruction Sets from Language Model Activations

提出PRISM方法，通过激活条件解码从冻结目标模型隐藏状态中恢复活跃指令集，利用法官引导的GRPO优化，在多种场景下优于基线方法。

当言语安全但行动致命：在隐藏状态风险空间中探究超越文本安全的物理危险

When Words Are Safe But Actions Kill: Probing Physical Jailbreak Beyond Textual Jailbreak in Hidden-State Risk Space

研究大语言模型中语言无害指令在现实世界中的物理危险与文本级内容危险是否相同，提出PRISM方法，通过隐藏状态方向分析等证明CD和PD可分离，PRISM在多个基准测试中表现良好，能检测物理危险而非仅依赖明确不安全措辞。

Logical Hidden Markov Models

Logical hidden Markov models (LOHMMs) upgrade traditional hidden Markov models to deal with sequences of structured symbols in the form of logical atoms, rather than flat characters. This note formally introduces LOHMMs and presents solutions to the three central inference problems for LOHMMs: evaluation, most likely hidden state sequence and parameter estimation. The resulting representation and algorithms are experimentally evaluated on problems from the domain of bioinformatics.

Second-Order Belief Hidden Markov Models

Hidden Markov Models (HMMs) are learning methods for pattern recognition. The probabilistic HMMs have been one of the most used techniques based on the Bayesian model. First-order probabilistic HMMs were adapted to the theory of belief functions such that Bayesian probabilities were replaced with mass functions. In this paper, we present a second-order Hidden Markov Model using belief functions. Previous works in belief HMMs have been focused on the first-order HMMs. We extend them to the second-order model.

Hidden Markov Models derived from Behavior Trees

Behavior trees are rapidly attracting interest in robotics and human task-related motion tracking. However no algorithms currently exist to track or identify parameters of BTs under noisy observations. We report a new relationship between BTs, augmented with statistical information, and Hidden Markov Models. Exploiting this relationship will allow application of many algorithms for HMMs (and dynamic Bayesian networks) to data acquired from BT-based systems.