SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1810.08575 · 2018

Supervising strong learners by amplifying weak experts

arXiv · PDF · Open in the Atlas

Code that ran

16 of 18 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 13 have no contract verdict; 3 failures of our instrument; 2 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/1810.08575.json · badge: /api/ran/1810.08575.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
paulfchristiano/amplification canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
rmoehn/amplification community (classifier label)
classifier basis not recorded on the link
13 ran · 2 no recorded run
FunctionStatusWhere it lives
add_logs Ran
ran on a synthesized input
rmoehn/amplification/amplification/tasks/graph.py
pointer only (licence: MIT) · get_code("1fea448f5339d7f6")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

bind Ran
ran; our contract draft was wrong, not the code
paulfchristiano/amplification/experiment.py
code served (permissive licence) · get_code("66c8330623902004")
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 3, 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 torch.nn as nn

torch.manual_seed(0)

# The bind function takes three parameters: var, val, and descriptor
# var: appears to be a variable identifier (string or similar)
# val: appears to be a value of any type
# descriptor: optional string descriptor, defaults to ''

# Looking at the function body, it just packages these into a nested tuple structure
# No specific attributes/methods are called on any arguments
# So we can use simple, minimal values

var = "my_variable"
val = 42
descriptor = "test_descriptor"

ARGS = {
    "var": var,
    "val": val,
    "descriptor": descriptor
}
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 list, ndim 3, 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
dfs Ran
ran on a synthesized input
rmoehn/amplification/amplification/tasks/equals.py
pointer only (licence: MIT) · get_code("ab64ff670a8d3fc4")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

flatten_first Ran
ran on a synthesized input
rmoehn/amplification/amplification/models/core.py
pointer only (licence: MIT) · get_code("78df87b7163affcc")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

flatten_to Ran
ran on a synthesized input
rmoehn/amplification/amplification/models/core.py
pointer only (licence: MIT) · get_code("582392fe7ea8743e")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

get_variable Ran
ran on a synthesized input
rmoehn/amplification/amplification/tf_utils.py
pointer only (licence: MIT) · get_code("dff55d08c6c73aca")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

ints Ran
ran on a synthesized input
rmoehn/amplification/amplification/models/asker.py
pointer only (licence: MIT) · get_code("4c6f302cb61b391d")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

ints_like Ran
ran on a synthesized input
rmoehn/amplification/amplification/models/asker.py
pointer only (licence: MIT) · get_code("3391fa2577a41f9b")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

is_context Ran
ran on a synthesized input
rmoehn/amplification/amplification/tf_utils.py
pointer only (licence: MIT) · get_code("3d51a440b6d573fd")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

is_known Ran
ran on a synthesized input
rmoehn/amplification/amplification/tasks/core.py
pointer only (licence: MIT) · get_code("d92ea87a218b4312")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

label Ran
ran; our contract draft was wrong, not the code
paulfchristiano/amplification/experiment.py
code served (permissive licence) · get_code("ad5aed415e724ebd")
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 3, 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

torch.manual_seed(0)

def bind(var, val, descriptor=''):
    return [((var, val, descriptor),)]

def label(descriptor):
    return bind(None, None, str(descriptor))

# The label function takes a single parameter: descriptor
# It converts descriptor to a string and passes it to bind as the third argument
# bind returns a list containing a tuple of ((var, val, descriptor),)

ARGS = {
    'descriptor': 'test_label'
}
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 list, ndim 3, 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
log Ran
ran on a synthesized input
rmoehn/amplification/amplification/tasks/graph.py
pointer only (licence: MIT) · get_code("d8d0370822941683")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

make_transcript Ran
ran on a synthesized input
rmoehn/amplification/amplification/models/asker.py
pointer only (licence: MIT) · get_code("3c22c9ecd14f3694")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

options Ran
ran; our contract draft was wrong, not the code
paulfchristiano/amplification/experiment.py
code served (permissive licence) · get_code("32a0883f9eed30c0")
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 3, 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)

def each(*xs):
    return [y for x in xs for y in x]

def bind(var, val, descriptor=''):
    return [((var, val, descriptor),)]

def options(var, opts):
    return each(*[bind(var, val, descriptor) for val, descriptor in opts])

# Set up arguments for options function
var = "learning_rate"
opts = [
    (0.001, "small_lr"),
    (0.01, "medium_lr"),
    (0.1, "large_lr")
]

ARGS = {
    "var": var,
    "opts": opts
}
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 3, 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
sample Ran
ran on a synthesized input
rmoehn/amplification/amplification/tasks/core.py
pointer only (licence: MIT) · get_code("4639fc0079859153")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

to_tags Ran
ran on a synthesized input
rmoehn/amplification/amplification/tf_utils.py
pointer only (licence: MIT) · get_code("18d8bdd503c65d78")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

multinomial Not yet run
unverified (harvested, no recorded run)
rmoehn/amplification/amplification/models/core.py
pointer only (licence: MIT) · get_code("cb49c8ca24a0f6f7")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

uniform Not yet run
unverified (harvested, no recorded run)
rmoehn/amplification/amplification/tasks/core.py
pointer only (licence: MIT) · get_code("c6ae75112108e1dd")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, rmoehn/amplification, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

Also present in other papers' code

Of 18 samples lifted from this paper's linked repositories, 18 carry a function identity and 0 do not. Of the 18 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 18 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

Many real world learning tasks involve complex or hard-to-specify objectives, and using an easier-to-specify proxy can lead to poor performance or misaligned behavior. One solution is to have humans provide a training signal by demonstrating or judging performance, but this approach fails if the task is too complicated for a human to directly evaluate. We propose Iterated Amplification, an alternative training strategy which progressively builds up a training signal for difficult problems by combining solutions to easier subproblems. Iterated Amplification is closely related to Expert Iteration (Anthony et al., 2017; Silver et al., 2017), except that it uses no external reward function. We present results in algorithmic environments, showing that Iterated Amplification can efficiently learn complex behaviors.

For agents

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

get_harvested_code_for_paper("1810.08575")
get_code_for_paper("1810.08575")
have("1810.08575")

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

curl https://syntology.ai/api/ran/1810.08575.json

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

[![Syntology run record](https://syntology.ai/api/ran/1810.08575.svg)](https://syntology.ai/paper/1810.08575)

Connect an agent — have() is free.