SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2511.08616 · ICLR · 2025

Reasoning on Time-Series for Financial Technical Analysis

Tat-Seng Chua, Kelvin Koa, Jan Chen, Yunshan Ma, Huanhuan Zheng

arXiv · PDF · Open in the Atlas

Code that ran

5 of 5 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; 3 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 2 failures of our instrument; 0 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/2511.08616.json · badge: /api/ran/2511.08616.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
chen-jan/VTA community
Papers with Code archive
5 ran · 0 no recorded run
FunctionStatusWhere it lives
format_array Ran
ran; our contract draft was wrong, not the code
chen-jan/VTA/grpo/grpo_utils.py
code served (permissive licence) · get_code("0ca488eca4e32ee8")
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 str. 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

# Test case 1: None input
arr_none = None

# Test case 2: List input
arr_list = [1.23456, 2.34567, 3.45678, 4.56789]

# Test case 3: NumPy array with floats
arr_numpy = np.array([[1.23456789, 2.3456789], [3.456789, 4.56789]])

# Test case 4: 1D array
arr_1d = np.array([0.0001234, 0.001234, 0.01234, 0.1234])

# Test case 5: 2D array with mixed scales
arr_2d = np.array([[1e-5, 1e-4, 1e-3], [1e2, 1e3, 1e4]])

# Use the simplest case for ARGS
ARGS = {
    'arr': np.array([1.23456, 2.34567, 3.45678])
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
format_input_stats Ran
ran; our contract draft was wrong, not the code
chen-jan/VTA/grpo/grpo_utils.py
code served (permissive licence) · get_code("d12beaa29aeb55a9")
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) and returned str. 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

# Create a minimal stats_row dictionary with all required fields
stats_row = {
    'sequence': '[100.5, 101.2, 99.8, 102.1, 100.9, 103.2, 102.5, 101.8, 104.1, 103.5]',
    'min': 99.8,
    'min_timestep': 2,
    'max': 104.1,
    'max_timestep': 8,
    'mean': 101.96,
    'median': 102.05,
    'median_timestep': 5,
    'sma': '[101.2, 101.5, 101.8, 102.1, 102.4]',
    'ema': '[100.8, 101.2, 101.6, 102.0, 102.4]',
    'mom': '[0.7, -1.4, 2.3, -1.2, 2.3]',
    'rsi': '[55.2, 48.1, 62.3, 45.5, 65.1]',
    'macd': '[0.15, 0.22, 0.18, 0.25, 0.20]',
    'willr': '[-25.5, -45.2, -15.8, -55.1, -10.3]',
    'cci': '[45.2, -12.5, 78.3, -35.1, 92.5]',
    'adx': '[22.5, 25.1, 28.3, 26.8, 30.2]',
    'bbands_upper': '[103.2, 103.8, 104.1, 104.5, 104.8]',
    'bbands_middle': '[101.5, 101.8, 102.1, 102.4, 102.7]',
    'bbands_lower': '[99.8, 100.1, 100.2, 100.3, 100.6]',
    'stoch_k': '[65.2, 48.1, 75.3, 42.5, 82.1]',
    'stoch_d': '[62.1, 55.2, 68.5, 52.3, 75.8]',
}

ARGS = {
    'stats_row': stats_row,
    'simplified': True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
safe_extract_float Ran
ran, honoured the contract we drafted
chen-jan/VTA/grpo/grpo_utils.py
code served (permissive licence) · get_code("b5b65fda0663b6e9")
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 float, ndim 0, dtype family float, shape () 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
import numpy as np

torch.manual_seed(0)

# The function safe_extract_float takes a single parameter 'value'
# It handles strings (including bracket-wrapped strings), ints, floats, and lists
# Let's create test cases for each branch

ARGS = {
    "value": "[1.23]"  # String with brackets - will extract 1.23
}
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 float, ndim 0, dtype family float, shape () — shape () 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
safe_extract_int Ran
ran, honoured the contract we drafted
chen-jan/VTA/grpo/grpo_utils.py
code served (permissive licence) · get_code("0937633e65a9ecee")
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 int, ndim 0, dtype family int, shape () 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
import numpy as np

torch.manual_seed(0)

ARGS = {
    "value": "[1.23]"
}
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 int, ndim 0, dtype family int, shape () — shape () 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
safe_parse_array Ran
ran, honoured the contract we drafted
chen-jan/VTA/grpo/grpo_utils.py
code served (permissive licence) · get_code("388eb35a6b3cbcce")
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.

import torch
import numpy as np

torch.manual_seed(0)

ARGS = {
    "value": "[1.5, 2.3, 3.7]"
}
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 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

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 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 5 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

While Large Language Models have been used to produce interpretable stock forecasts, they mainly focus on analyzing textual reports but not historical price data, also known as Technical Analysis. This task is challenging as it switches between domains: the stock price inputs and outputs lie in the time-series domain, while the reasoning step should be in natural language. In this work, we introduce Verbal Technical Analysis (VTA), a novel framework that combine verbal and latent reasoning to produce stock time-series forecasts that are both accurate and interpretable. To reason over time-series, we convert stock price data into textual annotations and optimize the reasoning trace using an inverse Mean Squared Error (MSE) reward objective. To produce time-series outputs from textual reasoning, we condition the outputs of a time-series backbone model on the reasoning-based attributes. Experiments on stock datasets across U.S., Chinese, and European markets show that VTA achieves state-of-the-art forecasting accuracy, while the reasoning traces also perform well on evaluation metrics judged by industry experts. Our code is available at: https://github.com/chen-jan/VTA.

For agents

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

get_harvested_code_for_paper("2511.08616")
get_code_for_paper("2511.08616")
have("2511.08616")

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

curl https://syntology.ai/api/ran/2511.08616.json

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

[![Syntology run record](https://syntology.ai/api/ran/2511.08616.svg)](https://syntology.ai/paper/2511.08616)

Connect an agent — have() is free.