SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.18603 · 2026

Enhancing Goal Inference via Correction Timing

Tesca Fitzgerald, Anjiabei Wang, Shuangge Wang

arXiv · PDF · Open in the Atlas

Code that ran

2 of 3 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; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 1 failure of our instrument; 1 no recorded run), measured 2026-11-22; 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/2602.18603.json · badge: /api/ran/2602.18603.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
iqr-lab/correction-timing canonical (classifier label)
classifier basis: census
2 ran · 1 no recorded run
FunctionStatusWhere it lives
load_data Ran
ran; our contract draft was wrong, not the code
iqr-lab/correction-timing/codes/where_model.py
code served (permissive licence) · get_code("f6992e3f43e7bbbd")
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 tuple. 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 pickle
import os
import tempfile

# Create temporary directories and mock pickle files
tmpdir = tempfile.mkdtemp()
splits_dir = os.path.join(tmpdir, 'splits')
features_dir = os.path.join(tmpdir, 'features')
os.makedirs(splits_dir, exist_ok=True)
os.makedirs(features_dir, exist_ok=True)

# Create mock split index files
target_nums = [0, 1, 2, 3]
shapes = ['circle', 'rectangle', 'triangle', 'square']

for target in target_nums:
    for shape in shapes:
        timing_ratio = 0.5
        split = 0.7
        rep = 1
        filename = os.path.join(splits_dir, f'indices_{int(100 * timing_ratio)}_{int(100 * split)}_shape_{shape}_target_{int(target)}_{int(rep)}.pkl')
        idx_dict = {
            'train': [0, 1],
            'val': [2],
            'test': [3]
        }
        with open(filename, 'wb') as f:
            pickle.dump(idx_dict, f)

# Create mock features file
timing_ratio = 0.5
features_file = os.path.join(features_dir, f'corrected_features_{int(100 * timing_ratio)}.pkl')
all_data_pre = []
for i in range(4):
    all_data_pre.append({
        'features': True,
        'correction_pose_list': [
            [0.0, 0.0, 0.0, 0.5, 0.5],
            [0.1, 0.1, 0.1, 0.5, 0.5],
            [0.2, 0.2, 0.2, 0.5, 0.5],
        ]
    })

with open(features_file, 'wb') as f:
    pickle.dump(all_data_pre, f)

# Patch the file paths in the module by modifying sys.path expectations
import sys
sys.path.insert(0, tmpdir)

# Create a mock module context where '../splits' and '../features' resolve correctly
original_open = open

def mock_open_func(filename, *args, **kwargs):
    if '../splits/' in filename:
        filename = filename.replace('../splits/', os.path.join(splits_dir, ''))
    elif '../features/' in filename:
        filename = filename.replace('../features/', os.path.join(features_dir, ''))
    return original_open(filename, *args, **kwargs)

import builtins
builtins.open = mock_open_func

ARGS = {
    'timing_ratio': 0.5,
    'split': 0.7,
    'rep': 1
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple 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
normalize Ran
ran, honoured the contract we drafted
iqr-lab/correction-timing/codes/where_model.py
code served (permissive licence) · get_code("81c0963ab80ed74c")
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 tuple. 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

torch.manual_seed(0)

# Create small numpy arrays for X and y
# X: shape (2, 4) - 2 samples, 4 features
# y: shape (2, 3) - 2 samples, 3 targets
X = np.random.randn(2, 4).astype(np.float32)
y = np.random.randn(2, 3).astype(np.float32)

ARGS = {
    "X": X,
    "y": y,
}
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 tuple — a tuple 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
filter_data Not yet run
unverified (harvested, no recorded run)
iqr-lab/correction-timing/codes/where_model.py
code served (permissive licence) · get_code("104b7e2ec9e26645")
How it ran

no invocation recorded.

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 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 3 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

Corrections offer a natural modality for people to provide feedback to a robot, by (i) intervening in the robot's behavior when they believe the robot is failing (or will fail) the task objectives and (ii) modifying the robot's behavior to successfully fulfill the task. Each correction offers information on what the robot should and should not do, where the corrected behavior is more aligned with task objectives than the original behavior. Most prior work on learning from corrections involves interpreting a correction as a new demonstration (consisting of the modified robot behavior), or a preference (for the modified trajectory compared to the robot's original behavior). However, this overlooks one essential element of the correction feedback, which is the human's decision to intervene in the robot's behavior in the first place. This decision can be influenced by multiple factors including the robot's task progress, alignment with human expectations, dynamics, motion legibility, and optimality. In this work, we investigate whether the timing of this decision can offer a useful signal for inferring these taskrelevant influences. In particular, we investigate three potential applications for this learning signal: (1) identifying features of a robot's motion that may prompt people to correct it, (2) quickly inferring the final goal of a human's correction based on the timing and initial direction of their correction motion, and (3) learning more precise constraints for task objectives. Our results indicate that correction timing results in improved learning for the first two of these applications. Overall, our work provides new insights on the value of correction timing as a signal for robot learning.

For agents

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

get_harvested_code_for_paper("2602.18603")
get_code_for_paper("2602.18603")
have("2602.18603")

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

curl https://syntology.ai/api/ran/2602.18603.json

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

[![Syntology run record](https://syntology.ai/api/ran/2602.18603.svg)](https://syntology.ai/paper/2602.18603)

Connect an agent — have() is free.