Stuart Russell, Sam Toyer, Rohin Shah, Andrew Critch
We lifted 10 functions out of this paper's own repositories and ran 9 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 |
|---|---|---|
| qxcv/magical | canonical | 9 of 10 |
| Function | Status | Where it lives |
|---|---|---|
| darken_rgb | Ran | qxcv/magical/magical/style.py code served (permissive licence) · get_code("6009fc6a277441d3") |
| get_unique_fn | Ran | qxcv/magical/magical/__main__.py code served (permissive licence) · get_code("affa8da28334940f") |
| latexify_results | Ran | qxcv/magical/magical/evaluation.py code served (permissive licence) · get_code("6d496932d2281447") |
| lighten_rgb | Ran | qxcv/magical/magical/style.py code served (permissive licence) · get_code("2d34512f6bbd54b5") |
| make_circle | Ran | qxcv/magical/magical/gym_render.py code served (permissive licence) · get_code("ca242cebfe737210") |
| regular_poly_apothem_to_side_legnth | Ran | qxcv/magical/magical/geom.py code served (permissive licence) · get_code("9995efb7b9dd9a25") |
| regular_poly_circ_rad_to_side_length | Ran | qxcv/magical/magical/geom.py code served (permissive licence) · get_code("042770b0f4e8a0c0") |
| regular_poly_circumrad | Ran | qxcv/magical/magical/geom.py code served (permissive licence) · get_code("c5ceb3580b3286f5") |
| rgb | Ran | qxcv/magical/magical/style.py code served (permissive licence) · get_code("4010fcbb54d3f347") |
| get_display | Not yet run | qxcv/magical/magical/gym_render.py code served (permissive licence) · get_code("9787f11078da23f3") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Imitation Learning (IL) algorithms are typically evaluated in the same environment that was used to create demonstrations. This rewards precise reproduction of demonstrations in one particular environment, but provides little information about how robustly an algorithm can generalise the demonstrator's intent to substantially different deployment settings. This paper presents the MAGICAL benchmark suite, which permits systematic evaluation of generalisation by quantifying robustness to different kinds of distribution shift that an IL algorithm is likely to encounter in practice. Using the MAGICAL suite, we confirm that existing IL algorithms overfit significantly to the context in which demonstrations are provided. We also show that standard methods for reducing overfitting are effective at creating narrow perceptual invariances, but are not sufficient to enable transfer to contexts that require substantially different behaviour, which suggests that new approaches will be needed in order to robustly generalise demonstrator intent. Code and data for the MAGICAL suite is available at https://github.com/qxcv/magical/.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2011.00401")
get_code_for_paper("2011.00401")
have("2011.00401")
Connect an agent — have() is free.