← Back to all news
arXiv AI

Same Behavior, Different Mechanism: What Finite Experiments Can Certify About Integrated Information

The paper investigates when the integrated information measure ζ (phi) can be reliably identified from finite, noisy experiments in small stochastic networks. It shows that even identical input‑output behavior can hide large differences in phi, that partial recordings impose hard precision limits, and that phi is highly sensitive to infinitesimal mechanism changes, often misreported by surrogate models. The authors prove that a lower bound on phi can be certified with controlled experiments and quantify the sample size needed for statistical confidence.

Related stories

arXiv Computation and Language
Sep 18

Message capacity and claim wording set the transition points of collective truth-finding in language-model networks

The study investigates how limited reading capacity and claim wording influence consensus outcomes in language‑model networks. By modeling message capacity as the number of messages an agent reads, the authors show that when agents read fewer than about 6.4 messages on average, a wrong consensus becomes unreachable. However, the wording of a claim—its inherent threshold—can override this effect, leading to incorrect consensus even when most agents start correct.

By Makoto Fukushima
llmsagentssafety
More like this →
arXiv Computation and Language
Sep 4

How Much Do Circuits Tell Us? Measuring the Consistency and Specificity of Language Model Circuits

The study investigates the consistency and specificity of language model circuits across six tasks and five models, focusing on component-level (attention heads and MLP blocks) and neuron-level circuits. Component-level circuits are highly consistent and causally important but lack task specificity, as ablating a circuit for one task similarly harms performance on other tasks. Neuron-level circuits show higher task specificity but lower consistency, with overlap mainly between closely related tasks. The analysis of Llama‑3.2‑3B reveals that shared components are predominantly MLP blocks, while attention heads act as generic attention‑sink heads.

By Michael Li, Nishant Subramani
llmssafety
More like this →
arXiv AI
Sep 10

Causal Attribution for Agentic Decisions: Estimators, Coupling, and a Traceability Specification

The paper presents a framework for causal attribution in agentic AI systems, outlining estimators and conditions where they fail. It distinguishes between marginal total effects and common‑random‑number total effects, introduces a natural direct effect under pinned downstreams, and derives a coupling method to keep direct effects estimable. The authors also propose a traceability specification to meet upcoming regulatory requirements for high‑risk AI systems.

By Ajay Pravin Mahale (Hochschule Trier)
agents
More like this →
arXiv Machine Learning
Oct 7

Finding the Heads and the Neurons Responsible for Network Information Retrieval in Language Models

The study investigates whether particular attention heads and individual neurons within those heads in language models are responsible for detecting network infrastructure information—specifically hostnames paired with IP addresses. Using causal ablation and selective testing across five models from three architecture families, the authors find that a small subset of heads reliably identifies such information with near-perfect accuracy. However, the extent to which this responsibility is concentrated in a single neuron varies by model; in some cases a single neuron suffices, while in others the signal is distributed across the head. The findings generalize to an independent reverse‑DNS dataset, though single‑neuron detectors are less robust.

By Abdul Kadir (University of Oldenburg, Oldenburg, Germany, German Research Center for Artificial Intelligence), Md Mohasin Hossain (German Research Center for Artificial Intelligence, Saarland University, Saarbrucken, Germany), Daniel Sonntag (University of Oldenburg, Oldenburg, Germany, German Research Center for Artificial Intelligence)
llmsbenchmarks
More like this →
arXiv AI
Sep 25

Every Component Is a Lookup: One Linear Graph for Interaction, Composition and Attribution

The paper proposes that two architectural assumptions—(1) attention and MLPs share a key‑value form <phi(S)>U, and (2) components read from an additive residual stream—are sufficient to answer three interpretability questions: component interaction, information routing, and token attribution. By treating these selections as a computational graph, the authors develop Unpack, a backward attribution method that validates interaction scores, recovered routes, and token attribution against established tests across models ranging from 160M to 6.9B parameters. The study also shows that contribution and causal effect can differ, with a recognizable signature in how components change when a task is removed.

By Po-Kai Chen, Aske Plaat, Niki van Stein
llmsbenchmarkssafety
More like this →