Branching Universes

arXiv:2603.18147 · hep-th, astro-ph.CO, gr-qc, hep-ph · Submitted 2026-03-18 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: Next we'll be talking about the paper "Branching Universes".

Jocelyn: The paper was written by Anamaria Hell and Tatsuya Daniel from Kavli IPMU (WPI) and UTIAS and The University of Tokyo and Center for Data-Driven Discovery and Trottier Space Institute and Department of Physics and McGill University.

Vera: Stay tuned as we take you through the paper and discuss its implications.

Title: Vera: We are looking at a fascinating new paper titled "Branching Universes" by Anamaria Hell from the University of Tokyo and Tatsuya Daniel from McGill University.

Jocelyn: That title sounds like something straight out of a science fiction novel, Vera.

Vera: It certainly does, but the math behind it is very grounded in actual physics.

Jocelyn: When you say "branching," are we talking about a multiverse where we could actually end up in another reality?

Subrahmanyan: Not exactly in the way movies depict it, Jocelyn.

Vera: Right, Subrahmanyan, they aren't proposing parallel dimensions you can walk into.

Subrahmanyan: Precisely, because what Hell and Daniel are suggesting is that our universe might just be one specific realization of a much broader set of possibilities allowed by gravity. They use these mathematical structures called spatially constrained vector fields to show how different "branches" or versions of the universe could exist.

Jocelyn: So, instead of one fixed set of rules for how gravity works, there are several possible sets?

Subrahmanyan: Exactly, and the specific branch we inhabit would be determined by certain fundamental constants.

Vera: And the most exciting part for an observer like me is that this isn't just theoretical fluff.

Jocelyn: How can we actually see if we are in one of these branches or just in a standard universe?

Subrahmanyan: That is the brilliant part, as it all comes down to how gravitational waves travel through space.

Vera: If the speed of those waves isn't exactly the speed of light, it would tell us which branch we are living in.

Jocelyn: That sounds like a massive implication for our current models of the cosmos.

Subrahmanyan: It really is, because it provides a way to test if General Relativity is the absolute final word or just one specific case of something much larger.

Vera: We are going to look at how they actually prove this in the next segment.

Summary: Vera: We’ve established that "Branching Universes" suggests we might be in one of many possible versions of reality, depending on the speed of gravitational waves.

Jocelyn: I'm trying to wrap my head around how they actually construct these branches without breaking physics as we know it.

Subrahmanyan: They achieve this by adding vector fields to the standard equations of gravity, but with a very clever twist.

Vera: They don't add any new types of waves, do they?

Subrahmanyan: No, and that is crucial for making the theory work. In many other modified gravity theories, you end up with extra "modes" or particles that would have been detected by our instruments already.

Jocelyn: So they've found a way to change the behavior of gravity while keeping the number of gravitational wave types exactly the same as what Einstein predicted?

Subrahmanyan: They have, and they show that this modification results in two distinct branches for a de Sitter universe. In one branch, gravitational waves travel at exactly the speed of light, which is what we see in our current observations.

Vera: But in the second branch, the speed shifts away from c.

Jocelyn: Could that shift be so small that our current instruments like LIGO can't even detect it yet?

Subrahmanyan: It could be incredibly tiny, which is why it's such a powerful prediction. The paper shows that if we ever measure a deviation in the speed of gravitational waves, even by a fraction, we would know immediately that we are in that second branch.

Vera: It’s like having a speedometer for the universe itself.

Jocelyn: That's an incredible way to frame it, Vera.

Subrahmanyan: And if the speed is exactly c, the theory still works perfectly and just collapses back into the standard General Relativity we are used to.

Vera: Let's talk about how this affects things like black holes in the next part of our discussion.

Improvements: Vera: We've been discussing how "Branching Universes" might change our understanding of gravitational wave speeds, but it goes much deeper than that.

Jocelyn: I was reading about their "stealth black holes," and that sounds like a very specific kind of solution.

Subrahmanyan: It is quite elegant, actually. A stealth black hole is one that looks exactly like a standard Schwarzschild black hole in terms of how it affects nearby matter and light.

Vera: So even if we are in a "branching" universe, our solar system tests wouldn't catch it because the gravity still feels normal?

Subrahmanyan: Exactly, which is why this theory is so hard to rule out. It passes the solar system experiments with ease because the modifications only become apparent in much more extreme cosmological scales or through high-precision gravitational wave measurements.

Jocelyn: They also mention how these vector fields could link up with matter in new ways, right?

Subrahmanyan: Yes, they propose a framework where matter is non-minimally coupled to these fields, which could potentially explain dark energy.

Vera: So this might not just be a way to test gravity, but a way to finally understand what's driving the acceleration of the universe.

Jocelyn: Could this also provide clues about dark matter?

Subrahmanyan: It certainly opens that door, as the vector fields could interact with dark matter in ways that standard models don't account for.

Vera: This paper seems to provide a massive toolkit for theorists and observers alike.

Jocelyn: It feels like we are standing on the edge of a much larger map of possibilities.

Conclusion: Vera: We have covered a lot of ground today, from the theoretical branches proposed by Hell and Daniel to the very real way we might test them with our telescopes.

Jocelyn: It really changes how you look at the sky when you realize we might just be one realization of many possible universes.

Subrahmanyan: The beauty is that this isn't just a wild guess; it's a mathematically consistent framework that is waiting for the data to confirm it.

Vera: If LIGO or LISA finds even a tiny shift in wave speed, the textbooks will be rewritten overnight.

Jocelyn: And if they don't, we at least know we are living in the most "standard" branch possible.

Subrahmanyan: Either way, this paper moves us forward by giving us a clear target for our next generation of observations.

Vera: Thank you all for joining us to discuss "Branching Universes."

Jocelyn: We'll see you next time when we tackle another fascinating new paper.

Subrahmanyan: Goodbye for now!

Anamaria Hell, Tatsuya Daniel

Kavli IPMU (WPI) · UTIAS · The University of Tokyo · Center for Data-Driven Discovery · Trottier Space Institute · Department of Physics · McGill University

hep-th, astro-ph.CO, gr-qc, hep-ph

Submitted: 2026-03-18

Updated: 2026-09-11

Comments: 23 pages

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 41/100

The gist: This paper proposes that our Universe is one realization among many possible branches, a hypothesis that can be observationally tested through the modified dispersion relation of gravitational waves.

Key concepts

Spatially constrained vector fields
These are mathematical structures added to gravity equations to allow for different "branches" of the universe. This approach modifies gravitational behavior without adding new types of waves or particles, ensuring the theory remains consistent with existing observations that have not detected extra modes.
Branching Universes
This theory suggests our reality is one specific realization of many possible versions allowed by gravity. One branch matches General Relativity, with gravitational waves traveling at the speed of light, while a second branch features waves traveling at a different speed.
Stealth black holes
These are theoretical objects that appear identical to standard Schwarzschild black holes in how they affect nearby matter and light. They are difficult to rule out because they pass solar system experiments, only showing differences at extreme cosmological scales or through precise gravitational wave measurements.

Terminology

Summary

This paper proposes that our Universe is one realization among many possible branches, a hypothesis that can be observationally tested through the modified dispersion relation of gravitational waves. By utilizing a framework of spatially constrained vector fields non-minimally coupled to gravity, the authors demonstrate how to modify gravity without introducing new propagating modes, such as additional scalar or vector polarizations.

The theoretical mechanism

This framework considers theories where vector fields are sourced purely by gravity through an action lacking standard kinetic terms. Unlike massive gravity or scalar-tensor theories that often introduce new propagating modes, this approach modifies gravity without adding additional scalar or vector polarizations. Instead, it produces tensor perturbations that differ from standard general relativity while only propagating two gravitational modes.

In de Sitter spacetime, the theory yields two distinct branches:

  1. A branch where A 0 = 0, which recovers standard GR and gives rise to unit speed of propagation for the GWs.

  2. A branch where A 0 not equal to 0, characterized by a non-vanishing value of the vector field and a modified GW propagation speed.

In flat spacetime, the temporal component A 0 can take an arbitrary value, which acts as a vacuum value that might be observationally detectable through gravitational wave propagation speeds. For the GWs to remain well-behaved, certain bounds on the coupling and mass must be satisfied.

Key solutions and stealth black holes

The theory admits several significant solutions, including stealth black holes. These are special types of black holes that cannot be distinguished from the standard GR solution at background order but admit non-trivial values for the background fields.

These solutions include:

  • Schwarzschild-like metrics where A 0 is non-zero, which still recover weak gravitational potentials and pass solar system experiments such as the deflection of starlight.

  • Multiple branches corresponding to different values of A 0 and A 1, where the spatial component A 1 can also be zero or non-vanishing.

Observational implications

The proposed theory has significant implications for dark energy, dark matter, and the cosmological constant problem. Because the GW propagation speed is modified universally when A 0 is nonzero, precision measurements can determine which branch the Universe occupies. Furthermore, the non-minimally driven vector fields open a new framework for matter coupling that allows for complex interactions between external matter and gravity.

The paper highlights several specific observational connections:

  • LVK observations: Constraints on the deviation of GW speed from light can be used to constrain xi and potentially constrain cosmology itself.

  • Fast Radio Bursts (FRBs): Time-delay measurements of strongly-lensed FRBs can probe gravitational slip, testing whether the theory's black hole solutions are observationally viable.

  • Dark Energy: The model allows for a more general set of solutions where A 0 evolves in time, potentially interpreting modified cosmology as some form of evolving dark energy. This may provide a further hint towards resolving the cosmological constant problem because, in certain branches, the value of the cosmological constant does not affect spacetime evolution.

Improvements for AI systems

1. Branching-Aware Bayesian Inference Engines

  • What it can do: Instead of assuming the standard General Relativity (GR) constraint where the gravitational wave speed c T = 1, these engines incorporate the parameterized dispersion relation c T squared = M Pl squared - xi over M Pl squared + xi directly into their likelihood functions. This allows for the real-time, high-precision estimation of the coupling constant xi and the vector field vacuum expectation value from LIGO/Virgo/KAGRA (LVK) and LISA strain data, enabling the AI to determine which branch of the universe we inhabit.

2. Multi-Messenger Temporal Correlation Transformers

  • What it can do: By utilizing attention mechanisms to cross-correlate gravitational wave (GW) arrival times with electromagnetic signals from Fast Radio Bursts (FRBs), this system can specifically calculate the gravitational slip parameter gamma psi/phi. It can distinguish between standard GR and the branching models described in Section IV B, identifying whether a lensed FRB signal is consistent with non-trivial A 0 background values or standard Schwarzschild geometry.

3. Non-Minimally Coupled Cosmological Generative Models (NCC-GANs)

  • What it can do: These generative adversarial networks replace standard CDM evolution equations with the modified Friedmann and conservation equations (Eq. 24 and 27) provided in the paper. This allows for the synthesis of high-fidelity mock datasets for large-scale structure (LSS) surveys that account for time-evolving A 0 components and non-minimal matter coupling, providing a robust training ground for AI models designed to analyze future SKA or Euclid data.

4. Dispersion-Sensitive Waveform Anomaly Detectors

  • What it can do: This deep learning architecture is specifically trained to detect sub-percent deviations in the chirp signal of gravitational waves caused by the modified dispersion relation (Eq. 11). Unlike standard templates that match signals against GR waveforms, this system identifies anomalous waveforms that exhibit frequency-dependent propagation shifts, allowing for the detection of a non-zero xi even when it is orders of magnitude smaller than the Planck scale.

Sources

Related papers