Extend the desiderata framework beyond world-state representations

Determine whether and how the paper’s information-theoretic framework of sensitivity, specificity, invariance, and functionality can be extended to apply to research on neural representations that do not primarily carry information about the actual state of the world (such as representations of goals, motor commands, mental simulations, planning, and imagery).

Background

The paper formalizes four desiderata—sensitivity, specificity, invariance, and functionality—for evaluating evidence that a neural response represents a feature, primarily focusing on representations that carry information about the state of the world (past, present, or future).

Section 5.7 acknowledges that not all representations fit this paradigm (e.g., goals/desires, motor commands, and mental simulations) and discusses how, in principle, the desiderata could be reinterpreted for such cases. Despite illustrative examples (e.g., planning-related neural sequences and uncertainty representations), the authors explicitly note that a more thorough account of extending the framework to these non-paradigmatic cases remains to be developed.

References

We leave it to future work to determine more thoroughly whether and how our framework can be extended to apply to research about representations that do not paradigmatically fall under our framework.

— Desiderata of evidence for representation in neuroscience  (2403.14046 - Pohl et al., 2024) in Section 5.7, Extending the framework to other representations

The present CFN is deliberately minimal, and its renewal remains representation sensitive. Future work should determine whether larger systems can extend this mechanism to increasingly abstract semantic transformations, intermediate reasoning states, and autonomously generated information.

— Beyond Episodic AI: Cognitive Field Networks for Biologically Inspired Persistent Cognition  (2609.16752 - Chae, 15 Sep 2026) in Section VI, Conclusion

Open Question: The Starting Point as Syntax or Semantics? When conceiving of a goal, the proposal to think of goals in a syntactic or semantic form means that we have a choice: do we begin by committing to a goal as a syntactic or a semantic object, first? And, how does our theory and analysis need to change depending on our choice of this starting point?

— The syntax and semantics of goals  (2609.19448 - Abel et al., 16 Sep 2026) in Discussion, subsection 'Open Question: The Starting Point as Syntax or Semantics?' (pp. 11–12)

Despite its fine-grained nature, certain attributes may remain uncovered, and we plan to further refine the taxonomy in the future.

— RobotEQ-Video: A Video-Centric Benchmark for Social Proactive Intelligence with World-State Taxonomy  (2609.21371 - Che et al., 18 Sep 2026) in Section 6, Limitations and Future Work

Decoding dynamic visual content from EEG has advanced rapidly for visual perception, yet it remains unclear which aspects of neural representations are preserved when people internally replay dynamic scenes.

— Replaying the Movie in Your Mind Decoding Dynamic Visual Perception and Recall from EEG  (2610.13120 - Fang et al., 9 Oct 2026) in Abstract; Section 1, Introduction