SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1909.04925 · 2019

How Does BERT Answer Questions? A Layer-Wise Analysis of Transformer Representations

arXiv · PDF · Open in the Atlas

Code that ran

1 of 2 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 0 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/1909.04925.json · badge: /api/ran/1909.04925.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
bvanaken/explain-BERT-QA canonical (classifier label)
classifier basis: census
1 ran · 1 no recorded run
FunctionStatusWhere it lives
convert_text_example_to_features Ran
ran on a synthesized input
bvanaken/explain-BERT-QA/visualization/data_utils.py
code served (permissive licence) · get_code("c4c28ae4f2098ccb")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned TextInputFeatures. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

# Create a minimal tokenizer that implements the required interface
class SimpleTokenizer:
    def __init__(self):
        # Vocabulary mapping
        self.vocab = {
            '[CLS]': 101,
            '[SEP]': 102,
            '[PAD]': 0,
            'hello': 7592,
            'world': 2088,
            'this': 2023,
            'is': 2003,
            'a': 1037,
            'test': 3231,
        }
        self.word_list = list(self.vocab.keys())
    
    def tokenize(self, text):
        # Simple whitespace tokenization
        words = text.lower().split()
        tokens = []
        for word in words:
            # Simple word matching (strip punctuation for this minimal example)
            clean_word = word.strip('.,!?;:')
            if clean_word in self.vocab:
                tokens.append(clean_word)
            else:
                tokens.append(clean_word)
        return tokens
    
    def convert_tokens_to_ids(self, tokens):
        ids = []
        for token in tokens:
            if token in self.vocab:
                ids.append(self.vocab[token])
            else:
                # Return a default ID for unknown tokens
                ids.append(1)
        return ids

# Create a minimal example object
class TextExample:
    def __init__(self, text):
        self.text = text

# Set up arguments
example = TextExample(text="hello world this is a test")
max_seq_length = 16
tokenizer = SimpleTokenizer()

ARGS = {
    'example': example,
    'max_seq_length': max_seq_length,
    'tokenizer': tokenizer,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type TextInputFeatures — a TextInputFeatures output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a TextInputFeatures output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
convert_qa_example_to_features Not yet run
unverified (harvested, no recorded run)
bvanaken/explain-BERT-QA/visualization/data_utils.py
code served (permissive licence) · get_code("4dda142f0c07fe7e")
How it ran

no invocation recorded.

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 2 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

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

Abstract

Bidirectional Encoder Representations from Transformers (BERT) reach state-of-the-art results in a variety of Natural Language Processing tasks. However, understanding of their internal functioning is still insufficient and unsatisfactory. In order to better understand BERT and other Transformer-based models, we present a layer-wise analysis of BERT's hidden states. Unlike previous research, which mainly focuses on explaining Transformer models by their attention weights, we argue that hidden states contain equally valuable information. Specifically, our analysis focuses on models fine-tuned on the task of Question Answering (QA) as an example of a complex downstream task. We inspect how QA models transform token vectors in order to find the correct answer. To this end, we apply a set of general and QA-specific probing tasks that reveal the information stored in each representation layer. Our qualitative analysis of hidden state visualizations provides additional insights into BERT's reasoning process. Our results show that the transformations within BERT go through phases that are related to traditional pipeline tasks. The system can therefore implicitly incorporate task-specific information into its token representations. Furthermore, our analysis reveals that fine-tuning has little impact on the models' semantic abilities and that prediction errors can be recognized in the vector representations of even early layers.

For agents

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

get_harvested_code_for_paper("1909.04925")
get_code_for_paper("1909.04925")
have("1909.04925")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/1909.04925.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/1909.04925.svg)](https://syntology.ai/paper/1909.04925)

Connect an agent — have() is free.