SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2112.01753 · 2021

Probing Linguistic Information For Logical Inference In Pre-trained Language Models

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 14 functions out of this paper's own repositories and ran 11 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.

RepositoryRoleRan
eric11eca/inference-information-probing canonical 11 of 14
FunctionStatusWhere it lives
batched_index_select Ran eric11eca/inference-information-probing/jiant/ext/allennlp.py
code served (permissive licence) · get_code("e445a3b7d6a58154")
count_freq Ran eric11eca/inference-information-probing/analyze.py
code served (permissive licence) · get_code("df8b734fdb77932a")
evaluate Ran eric11eca/inference-information-probing/inform_prob/trainer.py
code served (permissive licence) · get_code("349a73ed4ded5853")
file_len Ran eric11eca/inference-information-probing/inform_prob/util.py
code served (permissive licence) · get_code("e6bd0ba803f299b8")
flat_collate_fn Ran eric11eca/inference-information-probing/jiant/tasks/core.py
code served (permissive licence) · get_code("8f7619b2828926ea")
get_fasttext Ran eric11eca/inference-information-probing/inform_prob/process.py
code served (permissive licence) · get_code("59cfe607c72c35ba")
masked_softmax Ran eric11eca/inference-information-probing/jiant/ext/allennlp.py
code served (permissive licence) · get_code("16cde5ff8ae0a586")
metadata_collate_fn Ran eric11eca/inference-information-probing/jiant/tasks/core.py
code served (permissive licence) · get_code("61ffc5549a75fe3c")
read_data Ran eric11eca/inference-information-probing/inform_prob/util.py
code served (permissive licence) · get_code("0b2cc230b2f4d63a")
resolve_is_lower_case Ran eric11eca/inference-information-probing/jiant/shared/model_resolution.py
code served (permissive licence) · get_code("e17e83a5650fd73c")
resolve_warmup_steps Ran eric11eca/inference-information-probing/jiant/shared/model_setup.py
code served (permissive licence) · get_code("36c8a80131f2e20a")
eval_all Not yet run eric11eca/inference-information-probing/inform_prob/trainer.py
code served (permissive licence) · get_code("ad20a9d93a38e9cd")
get_type_token_ratio Not yet run eric11eca/inference-information-probing/analyze.py
code served (permissive licence) · get_code("fa314e7bf8386566")
weighted_sum Not yet run eric11eca/inference-information-probing/jiant/ext/allennlp.py
code served (permissive licence) · get_code("2f417d486bcad523")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Progress in pre-trained language models has led to a surge of impressive results on downstream tasks for natural language understanding. Recent work on probing pre-trained language models uncovered a wide range of linguistic properties encoded in their contextualized representations. However, it is unclear whether they encode semantic knowledge that is crucial to symbolic inference methods. We propose a methodology for probing linguistic information for logical inference in pre-trained language model representations. Our probing datasets cover a list of linguistic phenomena required by major symbolic inference systems. We find that (i) pre-trained language models do encode several types of linguistic information for inference, but there are also some types of information that are weakly encoded, (ii) language models can effectively learn missing linguistic information through fine-tuning. Overall, our findings provide insights into which aspects of linguistic information for logical inference do language models and their pre-training procedures capture. Moreover, we have demonstrated language models' potential as semantic and background knowledge bases for supporting symbolic inference methods.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2112.01753")
get_code_for_paper("2112.01753")
have("2112.01753")

Connect an agent — have() is free.