Greg Steeg, Aram Galstyan, Rob Brekelmans, Vaden Masrani, Frank Wood
We lifted 7 functions out of this paper's own repositories and ran 5 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| vmasrani/tvo_all_in | reimplementation | 5 of 7 |
| Function | Status | Where it lives |
|---|---|---|
| _moment_binary_search | Ran | vmasrani/tvo_all_in/src/models/updates.py code served (permissive licence) · get_code("c970b124ca982d5d") |
| exponentiate_and_normalize | Ran | vmasrani/tvo_all_in/src/models/updates.py code served (permissive licence) · get_code("7109bd8c08853514") |
| get_total_log_weight | Ran | vmasrani/tvo_all_in/src/models/updates.py code served (permissive licence) · get_code("20fc17e810e4eb41") |
| lognormexp | Ran | vmasrani/tvo_all_in/src/models/updates.py code served (permissive licence) · get_code("a62a8b0d64301ef6") |
| reshape_data | Ran | vmasrani/tvo_all_in/data/download_omniglot.py code served (permissive licence) · get_code("3c5f00f416eb718f") |
| calc_exp | Not yet run | vmasrani/tvo_all_in/src/models/updates.py code served (permissive licence) · get_code("9b0d14830128cf11") |
| moment_search | Not yet run | vmasrani/tvo_all_in/src/models/updates.py code served (permissive licence) · get_code("2a0edead5ea9fa28") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
The recently proposed Thermodynamic Variational Objective (TVO) leverages thermodynamic integration to provide a family of variational inference objectives, which both tighten and generalize the ubiquitous Evidence Lower Bound (ELBO). However, the tightness of TVO bounds was not previously known, an expensive grid search was used to choose a "schedule" of intermediate distributions, and model learning suffered with ostensibly tighter bounds. In this work, we propose an exponential family interpretation of the geometric mixture curve underlying the TVO and various path sampling methods, which allows us to characterize the gap in TVO likelihood bounds as a sum of KL divergences. We propose to choose intermediate distributions using equal spacing in the moment parameters of our exponential family, which matches grid search performance and allows the schedule to adaptively update over the course of training. Finally, we derive a doubly reparameterized gradient estimator which improves model learning and allows the TVO to benefit from more refined bounds. To further contextualize our contributions, we provide a unified framework for understanding thermodynamic integration and the TVO using Taylor series remainders.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2007.00642")
get_code_for_paper("2007.00642")
have("2007.00642")
Connect an agent — have() is free.