arXiv subjects

Jianfeng Xu

Publications and source records attributed to Jianfeng Xu.

At least 19 recordsLinked to original sources

arXiv2026

BridgeMem: Causal Dyadic Transition Residuals for Temporal Knowledge Graph Forecasting

Temporal knowledge graph forecasting aims to infer future relational facts from the temporal structure of observed events. Existing forecasters mainly summarize history through entity states, relation states, paths, or exact recurrence. These views often miss pair-specific transition evidence, that is, the way prior relations between the query actor and a candidate change the odds of the target relation. We introduce BridgeMem, which estimates this quantity as a residual added to the log scores of a frozen full-vocabulary forecaster. For each candidate, BridgeMem retrieves the pair's events that strictly precede t, encodes their relations, directions, and lags, and converts them into a likelihood-ratio correction. A support-adaptive empirical-Bayes reader trusts exact transition counts where they are abundant and backs off to a learned attention estimator where they are sparse. The backbone's own uncertainty gates the correction, so confident queries and candidates without dyadic history are left unchanged. On five benchmarks, BridgeMem improves on the strongest of nine baselines from 2021--2026 in all 20 filtered MRR and Hits@{1,3,10} comparisons, with MRR gains of 0.0213, 0.0164, 0.0216, 0.0112, and 0.0028 over the best prior result. These results show the value of explicit dyadic transition modeling.

cs.LG↗

arXiv2026

ALOE: Semantically Addressed Low-Rank Operators for Knowledge Editing

Knowledge editing changes what a model knows by modifying parameters so that a requested fact updates while unrelated behavior is preserved. This is usually treated as a write problem, but editing also involves an address problem: deciding which hidden states should receive the new residual. An update that activates too narrowly memorizes one prompt, while one that activates too broadly disrupts neighboring knowledge. Parametric editors encode this scope implicitly, whereas memory-based editors make the selection explicit but keep it outside the edited model. We propose ALOE (Addressed Low-rank Operator for Editing), which learns semantic addresses from paraphrases and hard same-subject negatives, aligns them with autoregressive hidden states through rollout refinement and gate calibration, and embeds the resulting gated low-rank operator within one MLP layer, so that the deployed model runs in a single forward pass with no external retriever or auxiliary router. Evaluated on CounterFact, ZSRE, and KnowEdit across three 7--8B model families, ALOE achieves efficacy between 0.955 and 0.999 and locality between 0.981 and 1.000; mechanistic analyses confirm that the learned geometry separates competing edits and that calibration suppresses out-of-scope activation. The remaining errors concentrate in paraphrase coverage and write fitting.

cs.AI↗

arXiv2026

Reasoning Instructions Can Break Answer Decoding in Vision--Language Models

Chain-of-thought (CoT) instructions can distort multiple-choice VLM evaluation when a scorer appends a reasoning cue but reads answer-label logits before the model generates any rationale. We call this CoT-prefix scoring. On ScienceQA, Qwen2.5-VL-7B drops from 80.76% to 45.48%, and across five option-content permutations 93.54% of CoT-prefix predictions select the first slot. Condition-matched linear probes recover 78.94% from the same hidden states, while free generation restores 75.24%, showing that the answer often survives the prefix and the immediate readout fails. Vocabulary and layer diagnostics explain the mismatch: probability mass moves toward continuation tokens, while answer information remains linearly accessible in late layers. The effect recurs with varying severity across datasets and models, though not universally. These results show that CoT-prefix scoring can confound model knowledge with an evaluation-interface mismatch and should be avoided unless the requested and scored output events are aligned.

cs.CL↗

arXiv2026

Proof-Valid Caching under Premise Erasures: Local Structural Limits and Shared-Workload Gains

We study reliable query recovery under independent premise erasures in semantically transparent caching systems, where every cached object must be a logical consequence of the premise base. Recovery succeeds only when the query remains derivable from surviving premises and the cache. Under a deterministic canonical-witness regime, we prove a query-local projection theorem and an exact residual-leaf law: recovery fails exactly when an erased base leaf retains a cache-free path to the query. Single-query design becomes weighted partial path interception. For shared workloads, we introduce semantic modules and derive exact reliability laws under joint and maximal-error criteria. The shared-module cache is exactly optimal under exact module routing and homogeneous costs, whereas optimal selection in general derivation DAGs is NP-complete at depth two. Against a coded benchmark recovering workload-relevant leaf payloads, MDS parity caching is optimal up to one packet. Leaf-only transparency incurs a first-order overhead inversely proportional to the erasure rate; shared modules multiply that inverse-erasure-rate scaling by the module-to-leaf cost ratio divided by the number of protected leaves. A Datalog instance and Monte Carlo checks illustrate the theory. For derivation-structured content, the results provide exact stochastic-erasure counterparts of function-correcting storage and an exact distributional quantification of maximal recoverability.

cs.IT↗

arXiv2026

Averaged Evaluation Masks Capability Trade-Offs: Multi-Source Calibration for High-Sparsity LLM Pruning

Calibration data are often treated as a minor implementation detail in post-training LLM pruning because averaged evaluations suggest only modest effects. We show that this conclusion is an averaging artifact: at 60\% SparseGPT sparsity, calibration strategies separated by only 2.85 points in averaged commonsense accuracy differ by 51.9 points in Code retention. Across 15 sources, capability-decomposed analysis reveals an opposing pattern: calibration perplexity is positively associated with General retention but negatively associated with Math or Code retention, leaving no evaluated single source uniformly strong across capabilities. This finding motivates capability-balanced multi-source calibration. Under the same calibration budget, a balanced real-data mixture outperforms every evaluated single source on LLaMA-3.1-8B, beating C4 by 18.8 points; the advantage grows with sparsity and persists on LLaMA-3.1-70B. Because the original pretraining data of advanced LLMs are often inaccessible, we further introduce Information-Guided Self-Calibration for Pruning (IGSP). Using only the base model and evaluation taxonomy, IGSP generates capability-stratified pools and selects low-redundancy samples within capability-specific perplexity ranges, outperforming Self-Cal and SGS by up to 4.8 points. Together, these results recast calibration as a capability-coverage problem and identify multi-source design as a practical principle for preserving capabilities in high-sparsity LLM pruning.

cs.LG↗

arXiv2026

Closure-Preserving Rate-Distortion for Reversible Logging

We study semantic compression of reversible-execution evidence for rollback reasoning. A run is a finite fact base; rollback semantics are modeled by a monotone closure operator induced by function-free Horn rules. A single edit replaces one fact by another; fidelity is the Jaccard discrepancy of the resulting closures, yielding a finite-alphabet distortion for rate-distortion analysis. A deterministic deletion scan decomposes the log into an irredundant core--preserving the closure--and a redundant remainder. Under admissible reconstructions (facts entailed by the original log), redundant facts are distortion-invisible, reducing the semantic rate-distortion function to a core-only optimization scaled by the core probability mass. At zero distortion, the optimal rate is a hypergraph entropy induced by overlaps of zero-distortion reconstruction sets on the core. We introduce a rollback-task loss based on a rollback observable, deriving parallel endpoint and factorization laws. The framework is instantiated on reversible causal nets and discussed in the event-structure view, showing how reversing disciplines yield different cores and compression frontiers. Numerical evaluation uses Blahut-Arimoto to design single-letter test channels and Monte Carlo reconstruction to assess end-to-end degradation at the log level.

cs.IT↗

arXiv2026

Rate-Distortion Theory for Deductive Sources under Closure Fidelity

We study lossy compression of a finite statement source generated in a fixed deductive environment. The source symbols are statements in a knowledge base endowed with a shared proof system, and reconstruction fidelity is measured by preservation of deductive closure rather than by symbolwise equality. Fixing the proof system and a canonical scan order yields a decomposition of the source alphabet into an irredundant core and redundant stored consequences. At zero distortion, each core symbol induces a set of distortion-free reconstructions. In the nonconfusable (disjoint-core) regime, we show that the minimum zero-distortion rate equals the source mass of the core times the entropy of the source conditioned on that core. In the general confusable-core regime, we characterise the exact zero-distortion rate via a hypergraph-entropy quantity induced by jointly realisable core subsets, with a reduction to Korner-style graph entropy under a natural pairwise realisability condition. For reconstruction alphabets contained in the deductive closure of the source knowledge base, we further prove that the full rate-distortion function depends only on the core, so redundant states are invisible to both rate and distortion. Finally, when the decoder is limited to a bounded inference-depth budget (a bounded number of iterations of the immediate-consequence operator), we obtain an exact rate-depth-distortion characterisation. Under an additional order-robustness assumption identifying the chosen core with the order-free essential set, this characterisation interpolates between classical symbolwise compression and unconstrained deductive compression.

cs.IT↗

arXiv2026

Semantic Rate-Distortion Theory: Deductive Compression and Closure Fidelity

Shannon's rate-distortion theory treats source symbols as unstructured labels. When the source is a knowledge base equipped with a logical proof system, a natural fidelity criterion is closure fidelity: a reconstruction is acceptable if it preserves the deductive closure of the original. This paper develops a rate-distortion theory under this criterion. Central to the theory is the irredundant core-a canonical generating set extracted by a fixed-order deletion procedure, from which the full deductive closure can be rederived. We prove that the zero-distortion semantic rate equals a quantity that is strictly below the classical entropy rate whenever the knowledge base contains redundant states. More generally, the full semantic rate-distortion function depends only on the core; redundant states are invisible to both rate and distortion. We derive a semantic source-channel separation theorem showing a semantic leverage phenomenon: under closure fidelity, the required source rate is reduced by an asymptotic leverage factor greater than one, allowing the same knowledge base to be communicated with proportionally fewer channel uses-not by violating Shannon capacity, but because redundant states become free. We also prove a strengthened Fano inequality that exploits core structure. For heterogeneous multi-agent communication, an overlap decomposition gives necessary and sufficient conditions for closure-reliable transmission and identifies a semantic bottleneck in broadcast settings that persists even over noiseless channels. All results are verified on Datalog instances with up to 24,000 base facts.

cs.IT↗

arXiv2026

Semantic Channel Theory: Deductive Compression and Structural Fidelity for Multi-Agent Communication

Shannon's information theory deliberately excludes message semantics. This paper develops a rigorous framework for semantic communication that integrates formal proof systems with Shannon-theoretic tools. We introduce an axiomatic information model comprising Lsem-definable state sets linked by computable enabling maps, and define the semantic channel as a composition of Markov kernels whose supports respect the enabling structure. A fixed proof system induces an irredundant semantic core and a derivation-depth stratification, enabling four distortion measures of increasing semantic depth: Hamming, closure, depth, and a parameterized composite. Six families of computable semantic channel invariants are defined and their inter-relationships established, including a data processing bound, a semantic Fano bound, and an ideal-channel collapse theorem. The central quantitative result is a deductive compression gain: under closure-based fidelity, the minimum block length is determined by the irredundant core size rather than the full knowledge-base size. We instantiate the framework for heterogeneous multi-agent communication, introducing an overlap decomposition that yields necessary and sufficient conditions for closure-reliable communication. A semantic bottleneck phenomenon is identified in broadcast settings: vocabulary mismatch imposes irreducible fidelity limitations even over noiseless carriers. All results are verified on an explicit Datalog instance.

cs.LO↗

arXiv2026

The Derivation Penalty in Premise-Erasure Caching: Capacity, Strong Converse, and Dispersion Dichotomy

We introduce an information-theoretic framework for caching in derivation-based reasoning engines under independent premise erasure. Two decoder models are compared: a coded scheme using an arbitrary bit-string cache with a general-purpose decoder, and a derivation-constrained scheme where the cache consists of logical facts and the decoder must produce a valid proof. Four coding theorems are established. The first proves that each derivation step carries a universal per-step information content determined by the base size. The second reveals an exponential capacity separation between linear-chain and balanced-merge Datalog architectures at equal depth. The third identifies a critical access frequency separating the regimes where caching and on-demand derivation are optimal. The fourth determines the minimum derivation-constrained cache under erasure, decomposing query information into reliable cache and noisy channel capacity. The central result is the derivation penalty: the ratio of the derivation-constrained cache to the coded cache converges to the reciprocal of the erasure rate, universally across query counts, overlap structures, and reliability targets. This penalty originates from a structural caching rigidity theorem showing that only cache facts within the target query's derivation DAG contribute to resilience, precluding cross-coordinate error correction. Beyond capacity, we prove a strong converse at the KL-divergence rate with Bahadur--Rao prefactors, a dispersion dichotomy (positive coded dispersion versus zero derivation-constrained dispersion), and a complete eight-regime phase diagram. The architecture-dependent depth-to-dependency mapping yields exponentially sharper phase transitions for the merge architecture. All results transfer across synonymous representations.

cs.IT↗

arXiv2026

Derivation Depth as an Information Metric: Axioms, Coding Theorems, and Storage--Computation Tradeoffs

We introduce derivation depth-a computable metric of the reasoning effort needed to answer a query based on a given set of premises. We model information as a two-layered structure linking abstract knowledge with physical carriers, and separate essential core facts from operational shortcuts. For any finite premise base, we define and prove the computability of derivation depth. By encoding reasoning traces and applying information-theoretic incompressibility arguments, we establish fundamental bounds linking depth to the descriptive complexity of queries. For frequently asked, information-rich queries, the minimal description length grows proportionally to depth times the logarithm of the knowledge base size. This leads to a practical storage-computation tradeoff: queries accessed beyond a critical threshold become cheaper to cache than recompute. We formulate optimal cache allocation as a mathematical optimization problem solvable with approximation guarantees and extend the framework to handle noisy or incomplete knowledge bases.

cs.IT↗

arXiv2026

DiffFace-Edit: A Diffusion-Based Facial Dataset for Forgery-Semantic Driven Deepfake Detection Analysis

Generative models now produce imperceptible, fine-grained manipulated faces, posing significant privacy risks. However, existing AI-generated face datasets generally lack focus on samples with fine-grained regional manipulations. Furthermore, no researchers have yet studied the real impact of splice attacks, which occur between real and manipulated samples, on detectors. We refer to these as detector-evasive samples. Based on this, we introduce the DiffFace-Edit dataset, which has the following advantages: 1) It contains over two million AI-generated fake images. 2) It features edits across eight facial regions (e.g., eyes, nose) and includes a richer variety of editing combinations, such as single-region and multi-region edits. Additionally, we specifically analyze the impact of detector-evasive samples on detection models. We conduct a comprehensive analysis of the dataset and propose a cross-domain evaluation that combines IMDL methods. Dataset will be available at https://github.com/ywh1093/DiffFace-Edit.

cs.CV↗

arXiv2025

Information Physics of Intelligence: Unifying Logical Depth and Entropy under Thermodynamic Constraints

The rapid scaling of artificial intelligence models has revealed a fundamental tension between model capacity (storage) and inference efficiency (computation). While classical information theory focuses on transmission and storage limits, it lacks a unified physical framework to quantify the thermodynamic costs of generating information from compressed laws versus retrieving it from memory. In this paper, we propose a theoretical framework that treats information processing as an enabling mapping from ontological states to carrier states. We introduce a novel metric, Derivation Entropy, which quantifies the effective work required to compute a target state from a given logical depth. By analyzing the interplay between Shannon entropy (storage) and computational complexity (time/energy), we demonstrate the existence of a critical phase transition point. Below this threshold, memory retrieval is thermodynamically favorable; above it, generative computation becomes the optimal strategy. This "Energy-Time-Space" conservation law provides a physical explanation for the efficiency of generative models and offers a rigorous mathematical bound for designing next-generation, energy-efficient AI architectures. Our findings suggest that the minimization of Derivation Entropy is a governing principle for the evolution of both biological and artificial intelligence.

cs.IT↗

arXiv2025

Information Science Principles of Machine Learning: A Causal Chain Meta-Framework Based on Formalized Information Mapping

This paper addresses the current lack of a unified formal framework in machine learning theory, as well as the absence of robust theoretical foundations for interpretability and ethical safety assurance. We first construct a formal information model, employing sets of well-formed formulas (WFFs) to explicitly define the ontological states and carrier mappings for the core components of machine learning. By introducing learnable and processable predicates, as well as learning and processing functions, we analyze the logical inference and constraint rules underlying causal chains in models, thereby establishing the Machine Learning Theory Meta-Framework (MLT-MF). Building upon this framework, we propose universal definitions for model interpretability and ethical safety, and rigorously prove and validate four key theorems: the equivalence between model interpretability and information existence, the constructive formulation of ethical safety assurance and two types of total variation distance (TVD) upper bounds. This work overcomes the limitations of previous fragmented approaches, providing a unified theoretical foundation from an information science perspective to systematically address the critical challenges currently facing machine learning.

cs.LO↗

arXiv2025

General Information Metrics for Improving AI Model Training Efficiency

To address the growing size of AI model training data and the lack of a universal data selection methodology-factors that significantly drive up training costs -- this paper presents the General Information Metrics Evaluation (GIME) method. GIME leverages general information metrics from Objective Information Theory (OIT), including volume, delay, scope, granularity, variety, duration, sampling rate, aggregation, coverage, distortion, and mismatch to optimize dataset selection for training purposes. Comprehensive experiments conducted across diverse domains, such as CTR Prediction, Civil Case Prediction, and Weather Forecasting, demonstrate that GIME effectively preserves model performance while substantially reducing both training time and costs. Additionally, applying GIME within the Judicial AI Program led to a remarkable 39.56% reduction in total model training expenses, underscoring its potential to support efficient and sustainable AI development.

cs.LG↗

arXiv2024

FairAdapter: Detecting AI-generated Images with Improved Fairness

The high-quality, realistic images generated by generative models pose significant challenges for exposing them.So far, data-driven deep neural networks have been justified as the most efficient forensics tools for the challenges. However, they may be over-fitted to certain semantics, resulting in considerable inconsistency in detection performance across different contents of generated samples. It could be regarded as an issue of detection fairness. In this paper, we propose a novel framework named Fairadapter to tackle the issue. In comparison with existing state-of-the-art methods, our model achieves improved fairness performance. Our project: https://github.com/AppleDogDog/FairnessDetection

cs.CV↗

arXiv2024

GOPT: Generalizable Online 3D Bin Packing via Transformer-based Deep Reinforcement Learning

Robotic object packing has broad practical applications in the logistics and automation industry, often formulated by researchers as the online 3D Bin Packing Problem (3D-BPP). However, existing DRL-based methods primarily focus on enhancing performance in limited packing environments while neglecting the ability to generalize across multiple environments characterized by different bin dimensions. To this end, we propose GOPT, a generalizable online 3D Bin Packing approach via Transformer-based deep reinforcement learning (DRL). First, we design a Placement Generator module to yield finite subspaces as placement candidates and the representation of the bin. Second, we propose a Packing Transformer, which fuses the features of the items and bin, to identify the spatial correlation between the item to be packed and available sub-spaces within the bin. Coupling these two components enables GOPT's ability to perform inference on bins of varying dimensions. We conduct extensive experiments and demonstrate that GOPT not only achieves superior performance against the baselines, but also exhibits excellent generalization capabilities. Furthermore, the deployment with a robot showcases the practical applicability of our method in the real world. The source code will be publicly available at https://github.com/Xiong5Heng/GOPT.

cs.RO↗

arXiv2022

Unification and Extension of Classic Information Principles

To formulate a universal framework of information theory is beneficial. This study proves that the sextuple model of the objective information theory (OIT) is a sufficient and necessary condition for discussing information with four basic postulations. It is demonstrated for each metric defined in the OIT, there is a corresponding example in classical information theories or commonly used principles. Furthermore, atomic information is defined as the indivisible elementary information and the volume additivity is proven for combinations of atomic information. Consequently, the information volume that a single quantum carrier can carry is derived and a theorem relating information volume to mass, energy, and time is proved. All these efforts illustrate that the OIT is a novel information theory that can unify a variety of classical information principles and even accurately reveal the quantitative relationship between information, matter and energy.

cs.IT↗