SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2410.20779 · 2024

Decoding Reading Goals from Eye Movements

arXiv · PDF · Open in the Atlas

Code that ran

8 of 9 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 0 have no contract verdict; 8 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/2410.20779.json · badge: /api/ran/2410.20779.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
lacclab/Goal-Decoding-from-Eye-Movements canonical (classifier label)
classifier basis: census
8 ran · 1 no recorded run
FunctionStatusWhere it lives
add_missing_features Ran
ran; our contract draft was wrong, not the code
lacclab/Goal-Decoding-from-Eye-Movements/src/ETDataset.py
code served (permissive licence) · get_code("6f8c4b735ae6e93c")
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 DataFrame. 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 numpy as np
import pandas as pd

# Create a small sample DataFrame with all required columns
np.random.seed(0)

n_rows = 10
ia_data = pd.DataFrame({
    'Reduced_POS': ['FUNC', 'NOUN', 'VERB', 'ADJ', 'UNKNOWN', 'FUNC', 'NOUN', 'VERB', 'ADJ', 'FUNC'],
    'Is_Content_Word': [False, True, True, True, False, False, True, True, True, False],
    'NEXT_FIX_INTEREST_AREA_INDEX': [2, 3, 1, 4, 2, 3, 1, 2, 5, 1],
    'CURRENT_FIX_INTEREST_AREA_INDEX': [1, 2, 3, 2, 1, 2, 2, 3, 4, 2],
    'IA_REGRESSION_IN_COUNT': [0, 1, 2, 1, 0, 1, 0, 2, 1, 1],
    'IA_REGRESSION_OUT_FULL_COUNT': [1, 0, 1, 0, 2, 1, 1, 0, 1, 0],
    'IA_FIXATION_COUNT': [3, 4, 5, 3, 4, 5, 3, 4, 5, 3],
    'Length': [1, 3, 6, 9, 2, 4, 7, 10, 3, 2],
    'IA_DWELL_TIME': [100.0, 150.0, 200.0, 175.0, 120.0, 160.0, 210.0, 190.0, 140.0, 130.0],
    'IA_FIRST_FIXATION_DURATION': [50.0, 75.0, 100.0, 80.0, 60.0, 85.0, 110.0, 95.0, 70.0, 65.0],
    'trial_id': [1, 1, 1, 1, 1, 2, 2, 2, 2, 2],
})

trial_groupby_columns = ['trial_id']

ARGS = {
    'ia_data': ia_data,
    'trial_groupby_columns': trial_groupby_columns,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataFrame — a DataFrame output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataFrame output the recorded ndim 2 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
compute_questions_cls_encodings Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
lacclab/Goal-Decoding-from-Eye-Movements/src/external_features_extractor.py
code served (permissive licence) · get_code("bf3ed24055344266")
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 Tensor, ndim 3, dtype family float, shape (2, 3, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape 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
from transformers import RobertaModel
from enum import StrEnum

class ReturnedTokensModes(StrEnum):
    CLS = 'cls'
    ALL_TOKENS = 'all_tokens'
    ALL_EXCEPT_CLS = 'all_except_cls'

torch.manual_seed(0)

# Create a minimal RobertaModel stand-in that matches the function's usage
class MinimalRobertaModel(torch.nn.Module):
    def __init__(self, hidden_size=4):
        super().__init__()
        self.hidden_size = hidden_size
        self.embedding = torch.nn.Embedding(1000, hidden_size)
        
    def forward(self, input_ids, attention_mask):
        # Return an object with last_hidden_state attribute
        batch_size, seq_len = input_ids.shape
        last_hidden_state = self.embedding(input_ids)
        
        class Output:
            pass
        
        output = Output()
        output.last_hidden_state = last_hidden_state
        return output
    
    def eval(self):
        return super().eval()

bert_encoder = MinimalRobertaModel(hidden_size=4)
bert_encoder.eval()

# Create test tensors matching the expected shapes
# batch_size=2, no_questions=3, max_q_len=5
batch_size = 2
no_questions = 3
max_q_len = 5

question_ids = torch.randint(0, 1000, (batch_size, no_questions, max_q_len))
question_masks = torch.ones((batch_size, no_questions, max_q_len), dtype=torch.long)

ARGS = {
    'bert_encoder': bert_encoder,
    'question_ids': question_ids,
    'question_masks': question_masks,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 3, 4) — shape (2, 3, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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_string_to_list Ran
ran; our contract draft was wrong, not the code
lacclab/Goal-Decoding-from-Eye-Movements/src/extract_last_layer_embeddings_eval_probe.py
code served (permissive licence) · get_code("93ed9e20175a7d7b")
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 list. 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 ast
import pandas as pd

# Create a pandas Series with string representations of lists
s = pd.Series([
    "[1, 2, 3]",
    "[4, 5, 6]",
    "['a', 'b']",
    "[True, False, None]"
])

ARGS = {
    "s": s
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list output has no array shape of its own; none is 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
create_and_configure_logger Ran
ran; our contract draft was wrong, not the code
lacclab/Goal-Decoding-from-Eye-Movements/src/extract_last_layer_embeddings_eval_probe.py
code served (permissive licence) · get_code("8f1f8fffd64e6e28")
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 Logger. 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 logging
import tempfile
import os

# Create a temporary directory for the log file
temp_dir = tempfile.gettempdir()
log_file = os.path.join(temp_dir, 'test_log.log')

ARGS = {
    'log_name': log_file
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Logger — a Logger output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Logger 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
defaultdict_to_df Ran
ran; our contract draft was wrong, not the code
lacclab/Goal-Decoding-from-Eye-Movements/src/extract_last_layer_embeddings_eval_probe.py
code served (permissive licence) · get_code("930c6279af006d50")
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 DataFrame. 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.

from collections import defaultdict
import numpy as np
import pandas as pd

# Create sample AUROC data
# Each value should be an array-like that np.mean and np.std can process
macro_auroc = defaultdict(list)
macro_auroc['eval_type_1'] = np.array([0.85, 0.87, 0.86])
macro_auroc['eval_type_2'] = np.array([0.90, 0.92, 0.91])

binary_auroc = defaultdict(list)
binary_auroc['eval_type_1'] = np.array([0.78, 0.80, 0.79])
binary_auroc['eval_type_2'] = np.array([0.88, 0.89, 0.87])

ARGS = {
    'macro_auroc': macro_auroc,
    'binary_auroc': binary_auroc,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataFrame — a DataFrame output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataFrame output the recorded ndim 2 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
encoding Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
lacclab/Goal-Decoding-from-Eye-Movements/src/external_features_extractor.py
code served (permissive licence) · get_code("08e943fa1b534ef5")
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 Tensor, ndim 2, dtype family float, shape (2, 8) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape 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
from transformers import RobertaModel
from enum import StrEnum

class ReturnedTokensModes(StrEnum):
    CLS = 'cls'
    ALL_TOKENS = 'all_tokens'
    ALL_EXCEPT_CLS = 'all_except_cls'

torch.manual_seed(0)

# Create a minimal RobertaModel
# RobertaModel requires a config object
from transformers import RobertaConfig

config = RobertaConfig(
    vocab_size=100,
    hidden_size=8,
    num_hidden_layers=2,
    num_attention_heads=2,
    intermediate_size=16,
    max_position_embeddings=16,
)

bert_encoder = RobertaModel(config)
bert_encoder.eval()

# Create input tensors
batch_size = 2
seq_length = 4

input_ids = torch.randint(0, 100, (batch_size, seq_length))
attention_mask = torch.ones((batch_size, seq_length), dtype=torch.long)

returned_tokens = ReturnedTokensModes.CLS

ARGS = {
    'bert_encoder': bert_encoder,
    'input_ids': input_ids,
    'attention_mask': attention_mask,
    'returned_tokens': returned_tokens,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 8) — shape (2, 8) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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
setup_callbacks Ran
ran; our contract draft was wrong, not the code
lacclab/Goal-Decoding-from-Eye-Movements/src/train_utils.py
code served (permissive licence) · get_code("53df8ab3fd95416b")
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 list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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.

from datetime import timedelta
from pathlib import Path
import tempfile

# Create a temporary directory for the checkpoint callback
temp_dir = tempfile.mkdtemp()
dir_path = Path(temp_dir)

# Define arguments
ARGS = {
    'dir_path': dir_path,
    'early_stopping_patience': 3,
    'max_time': timedelta(hours=1)
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
lightning — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
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
setup_logger Ran
ran; our contract draft was wrong, not the code
lacclab/Goal-Decoding-from-Eye-Movements/src/train_utils.py
code served (permissive licence) · get_code("23f4a555a4ea9dcf")
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 WandbLogger. 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 torch.nn as nn
from lightning.pytorch.loggers.wandb import WandbLogger

torch.manual_seed(0)

ARGS = {
    "wandb_project": "test_project",
    "wandb_entity": "test_entity",
    "wandb_job_type": "train"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
lightning, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type WandbLogger — a WandbLogger output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a WandbLogger 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
move_to_device Not yet run
unverified (harvested, no recorded run)
lacclab/Goal-Decoding-from-Eye-Movements/src/external_features_extractor.py
code served (permissive licence) · get_code("8a0aa5974a9b40a1")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 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 9 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

Readers can have different goals with respect to the text that they are reading. Can these goals be decoded from their eye movements over the text? In this work, we examine for the first time whether it is possible to distinguish between two types of common reading goals: information seeking and ordinary reading for comprehension. Using large-scale eye tracking data, we address this task with a wide range of models that cover different architectural and data representation strategies, and further introduce a new model ensemble. We find that transformer-based models with scanpath representations coupled with language modeling solve it most successfully, and that accurate predictions can be made in real time, long before the participant finished reading the text. We further introduce a new method for model performance analysis based on mixed effect modeling. Combining this method with rich textual annotations reveals key properties of textual items and participants that contribute to the difficulty of the task, and improves our understanding of the variability in eye movement patterns across the two reading regimes.

For agents

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

get_harvested_code_for_paper("2410.20779")
get_code_for_paper("2410.20779")
have("2410.20779")

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

curl https://syntology.ai/api/ran/2410.20779.json

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

[![Syntology run record](https://syntology.ai/api/ran/2410.20779.svg)](https://syntology.ai/paper/2410.20779)

Connect an agent — have() is free.