SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1606.00739 · 2016

Stochastic Structured Prediction under Bandit Feedback

arXiv · PDF · Open in the Atlas

Code that ran

11 of 14 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 11 have no contract verdict; 0 failures of our instrument; 3 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/1606.00739.json · badge: /api/ran/1606.00739.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
juliakreutzer/bandit-cdec community (classifier label)
classifier basis not recorded on the link
11 ran · 3 no recorded run
FunctionStatusWhere it lives
CountEF Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/python/cdec/sa/features.py
code served (permissive licence) · get_code("dfbdc79701875633")
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 math

class Context:
    def __init__(self, online=None, paircount=0):
        self.online = online
        self.paircount = paircount

# Create a context with no online component
ctx_offline = Context(online=None, paircount=5)

# Create a context with an online component
online_ctx = Context(online=None, paircount=3)
ctx_online = Context(online=online_ctx, paircount=7)

# Use the offline context as the main test case
ARGS = {
    "ctx": ctx_offline
}
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 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
EgivenF Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/python/cdec/sa/features.py
code served (permissive licence) · get_code("0a7d29cd68ed2cb1")
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 math

class MockContext:
    def __init__(self, online=None, paircount=10, fcount=100):
        self.online = online
        self.paircount = paircount
        self.fcount = fcount

# Create a context with online=None
ctx_offline = MockContext(online=None, paircount=10, fcount=100)

# Create a context with online component
ctx_online_component = MockContext(online=None, paircount=5, fcount=50)
ctx_online = MockContext(online=ctx_online_component, paircount=10, fcount=100)

ARGS = {
    "ctx": ctx_offline
}
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 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
SampleCountF Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/python/cdec/sa/features.py
code served (permissive licence) · get_code("a23f7b78a084fc18")
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 math

class MockContext:
    def __init__(self, online=None, fsample_count=0):
        self.online = online
        self.fsample_count = fsample_count

# Create a context without online component
ctx_offline = MockContext(online=None, fsample_count=5)

# Create a context with online component
ctx_with_online = MockContext(
    online=MockContext(online=None, fsample_count=3),
    fsample_count=5
)

# Test with offline context
ARGS = {
    "ctx": ctx_offline
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
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
escape Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/corpus/xml-tok.py
code served (permissive licence) · get_code("1ea28f9c5a22c5be")
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 torch

def escape(s):
    return s.replace('&', '&amp;').replace('>', '&gt;').replace('<', '&lt;').replace('"', '&quot;').replace("'", '&apos;')

ARGS = {
    's': "Hello & goodbye <world> \"quoted\" 'apostrophe'"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — 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
get_score_multilex Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/python/cdec/sa/online.py
code served (permissive licence) · get_code("49a14dfde374a61d")
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.

class BiLex:
    def __init__(self, f_dict, e_dict, fe_dict):
        self.f = f_dict
        self.e = e_dict
        self.fe = fe_dict

bilex1 = BiLex(
    f_dict={"word1": 5, "word2": 3},
    e_dict={"eword1": 4, "eword2": 6},
    fe_dict={("word1", "eword1"): 2, ("word2", "eword2"): 1}
)

bilex2 = BiLex(
    f_dict={"word1": 2, "word3": 4},
    e_dict={"eword1": 3, "eword3": 5},
    fe_dict={("word1", "eword1"): 1, ("word3", "eword3"): 2}
)

ARGS = {
    "f": "word1",
    "e": "eword1",
    "dir": 0,
    "bilex_list": [bilex1, bilex2]
}
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 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
gzopen Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/corpus/sample-dev-sets.py
code served (permissive licence) · get_code("ff14858de86a696a")
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 TextIOWrapper. 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 gzip
import tempfile
import os

# Create a temporary regular text file
temp_text_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_text_file.write("test content")
temp_text_file.close()

# Create a temporary gzipped file
temp_gz_file = tempfile.NamedTemporaryFile(delete=False, suffix='.gz')
temp_gz_file.close()
with gzip.open(temp_gz_file.name, 'wb') as f:
    f.write(b"gzipped test content")

ARGS = {
    'f': temp_text_file.name
}
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 TextIOWrapper — a TextIOWrapper 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
gzopen Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/corpus/tokenize-parallel.py
code served (permissive licence) · get_code("2142ad9aacfc4876")
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 TextIOWrapper. 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 gzip
import tempfile
import os

# Create a temporary file to use as the argument
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
temp_file.write("test content")
temp_file.close()
temp_file_path = temp_file.name

# Construct the ARGS dictionary
ARGS = {
    'f': temp_file_path
}
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 TextIOWrapper — a TextIOWrapper 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
match_utf8 Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/python/cdec/configobj.py
code served (permissive licence) · get_code("e60b9d981ac67440")
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 bool, 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

ARGS = {
    'encoding': 'utf-8'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type bool, 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
unescape Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/corpus/xml-tok.py
code served (permissive licence) · get_code("0c5b87b5a007d85c")
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 torch

ARGS = {
    's': '&gt;hello&lt;world&quot;test&apos;data&amp;more'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — 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
wc Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/corpus/sample-dev-sets.py
code served (permissive licence) · get_code("a2f4fb7985379b98")
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 tempfile
import os

# Create a temporary text file for testing
with tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt') as tmp:
    tmp.write("line1\n")
    tmp.write("line2\n")
    tmp.write("line3\n")
    tmp_path = tmp.name

try:
    # Define ARGS mapping the parameter 'f' to a valid file path
    ARGS = {
        'f': tmp_path
    }
finally:
    # Clean up the temporary file after defining ARGS
    pass
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 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
wc Ran
ran on a synthesized input
juliakreutzer/bandit-cdec/corpus/tokenize-parallel.py
code served (permissive licence) · get_code("ff0ff0b5ee67f98b")
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 tempfile
import os

# Create a temporary text file for testing
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test.txt")

# Write some lines to the file
with open(temp_file, 'w') as f:
    f.write("line 1\n")
    f.write("line 2\n")
    f.write("line 3\n")

ARGS = {
    'f': temp_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 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
extract Not yet run
unverified (harvested, no recorded run)
juliakreutzer/bandit-cdec/python/cdec/sa/extract.py
code served (permissive licence) · get_code("000822bea2505eb1")
How it ran

no invocation recorded.

getObj Not yet run
unverified (harvested, no recorded run)
juliakreutzer/bandit-cdec/python/cdec/configobj.py
code served (permissive licence) · get_code("42d0209877e4ef8b")
How it ran

no invocation recorded.

unrepr Not yet run
unverified (harvested, no recorded run)
juliakreutzer/bandit-cdec/python/cdec/configobj.py
code served (permissive licence) · get_code("470bad2afbb65f1c")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 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 14 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

Stochastic structured prediction under bandit feedback follows a learning protocol where on each of a sequence of iterations, the learner receives an input, predicts an output structure, and receives partial feedback in form of a task loss evaluation of the predicted structure. We present applications of this learning scenario to convex and non-convex objectives for structured prediction and analyze them as stochastic first-order methods. We present an experimental evaluation on problems of natural language processing over exponential output spaces, and compare convergence speed across different objectives under the practical criterion of optimal task performance on development data and the optimization-theoretic criterion of minimal squared gradient norm. Best results under both criteria are obtained for a non-convex objective for pairwise preference learning under bandit feedback.

For agents

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

get_harvested_code_for_paper("1606.00739")
get_code_for_paper("1606.00739")
have("1606.00739")

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

curl https://syntology.ai/api/ran/1606.00739.json

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

[![Syntology run record](https://syntology.ai/api/ran/1606.00739.svg)](https://syntology.ai/paper/1606.00739)

Connect an agent — have() is free.