SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2310.15047 · 2023

Implicit meta-learning may lead language models to trust more reliable sources

arXiv · PDF · Open in the Atlas

Code that ran

15 of 22 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, 2 violated it and 0 have no contract verdict; 10 failures of our instrument; 7 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/2310.15047.json · badge: /api/ran/2310.15047.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
krasheninnikov/internalization canonical (classifier label)
classifier basis: census
15 ran · 7 no recorded run
FunctionStatusWhere it lives
calculate_metrics Ran
ran; our contract draft was wrong, not the code
krasheninnikov/internalization/src/fewshot.py
pointer only (licence: NONE) · get_code("27effc0e19a1a9df")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

compute_exact_match Ran
ran, honoured the contract we drafted
krasheninnikov/internalization/src/metrics.py
pointer only (licence: NONE) · get_code("6e67afebffc30110")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

compute_f1 Ran
ran, honoured the contract we drafted
krasheninnikov/internalization/src/metrics.py
pointer only (licence: NONE) · get_code("e0d035eca2e2c3c5")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

create_tokenizer Ran
ran; our contract draft was wrong, not the code
krasheninnikov/internalization/src/lm_training_utils.py
pointer only (licence: NONE) · get_code("a7441b30297b1a30")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

finetune_single_stage Ran
ran; our contract draft was wrong, not the code
krasheninnikov/internalization/src/finetune_openai_models.py
pointer only (licence: NONE) · get_code("a9c947d467f63086")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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_cifar_loaders Ran
ran; our contract draft was wrong, not the code
krasheninnikov/internalization/vision_data_order_centroids.py
pointer only (licence: NONE) · get_code("2fb60475311c9a27")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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_cifar_loaders Ran
ran; our contract draft was wrong, not the code
krasheninnikov/internalization/vision_data_order_probing.py
pointer only (licence: NONE) · get_code("9ffbea5e2117e6d0")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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_gradient Ran
ran, honoured the contract we drafted
krasheninnikov/internalization/src/callbacks.py
pointer only (licence: NONE) · get_code("1bfb4e780e3caa9b")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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_orthonormal Ran
ran, violated the contract we drafted
krasheninnikov/internalization/activation_centroid_plotting.py
pointer only (licence: NONE) · get_code("f3a12c425050b2da")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

load_centroids_with_meta Ran
ran; our contract draft was wrong, not the code
krasheninnikov/internalization/centroid_3d_plot.py
pointer only (licence: NONE) · get_code("c3b0c173c5bbf5bb")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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_transforms Ran
ran; our contract draft was wrong, not the code
krasheninnikov/internalization/vision_data_order_centroids.py
pointer only (licence: NONE) · get_code("529a48b4dc68d507")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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_transforms Ran
ran; our contract draft was wrong, not the code
krasheninnikov/internalization/vision_data_order_probing.py
pointer only (licence: NONE) · get_code("257819637d0ed3e1")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

normalize_rows Ran
ran, violated the contract we drafted
krasheninnikov/internalization/activation_centroid_plotting.py
pointer only (licence: NONE) · get_code("6918fbf0270593a0")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

normalize_text Ran
ran; our contract draft was wrong, not the code
krasheninnikov/internalization/src/metrics.py
pointer only (licence: NONE) · get_code("12cb4aef8c35072c")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

pick_last_layer_and_token Ran
ran; our contract draft was wrong, not the code
krasheninnikov/internalization/activation_centroid_plotting.py
pointer only (licence: NONE) · get_code("bfa5ab71ffbb1027")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

finetune_two_stage Not yet run
unverified (harvested, no recorded run)
krasheninnikov/internalization/src/finetune_openai_models.py
pointer only (licence: NONE) · get_code("64b0420658254f36")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

generate_eval_prompts Not yet run
unverified (harvested, no recorded run)
krasheninnikov/internalization/src/fewshot.py
pointer only (licence: NONE) · get_code("5e249dd61a47ece2")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

get_imagenet32_loaders Not yet run
unverified (harvested, no recorded run)
krasheninnikov/internalization/vision_data_order_centroids.py
pointer only (licence: NONE) · get_code("c7bf6a8fca209c41")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

get_imagenet32_loaders Not yet run
unverified (harvested, no recorded run)
krasheninnikov/internalization/vision_data_order_probing.py
pointer only (licence: NONE) · get_code("84faca3d9d1c14d2")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

linear_probe Not yet run
unverified (harvested, no recorded run)
krasheninnikov/internalization/src/lm_training_utils.py
pointer only (licence: NONE) · get_code("fb9f795b86932411")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

pick_last_layer_and_token Not yet run
unverified (harvested, no recorded run)
krasheninnikov/internalization/centroid_3d_plot.py
pointer only (licence: NONE) · get_code("6d4a2884e219feeb")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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.

prompt_and_eval_hf_batched Not yet run
unverified (harvested, no recorded run)
krasheninnikov/internalization/src/fewshot.py
pointer only (licence: NONE) · get_code("63a902c0e9b81289")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, krasheninnikov/internalization when it was checked. 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 22 samples lifted from this paper's linked repositories, 22 carry a function identity and 0 do not. Of the 19 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 0 later) and 17 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.

FunctionAlso present inEarliest-datedOthers named
compute_exact_match 3 other papers
3 dated earlier
2110.07803 (2021-10)
Attacking Open-domain Question Answering by Injecting Misinformation
2304.13007 (2023-04)
Answering Questions by Meta-Reasoning over Multiple Chains of Thought
2305.12295 (2023-05)
Logic-LM: Empowering Large Language Models with Symbolic Solvers for Faithful Logical Reasoning
compute_f1 2 other papers
2 dated earlier
2304.13007 (2023-04)
Answering Questions by Meta-Reasoning over Multiple Chains of Thought
2305.12295 (2023-05)
Logic-LM: Empowering Large Language Models with Symbolic Solvers for Faithful Logical Reasoning

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

We demonstrate that LLMs may learn indicators of document usefulness and modulate their updates accordingly. We introduce random strings ("tags") as indicators of usefulness in a synthetic fine-tuning dataset. Fine-tuning on this dataset leads to implicit meta-learning (IML): in further fine-tuning, the model updates to make more use of text that is tagged as useful. We perform a thorough empirical investigation of this phenomenon, finding (among other things) that (i) it occurs in both pretrained LLMs and those trained from scratch, as well as on a vision task, and (ii) larger models and smaller batch sizes tend to give more IML. We also use probing to examine how IML changes the way models store knowledge in their parameters. Finally, we reflect on what our results might imply about capabilities, risks, and controllability of future AI systems. Our code can be found at https://github.com/krasheninnikov/internalization.

For agents

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

get_harvested_code_for_paper("2310.15047")
get_code_for_paper("2310.15047")
have("2310.15047")

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

curl https://syntology.ai/api/ran/2310.15047.json

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

[![Syntology run record](https://syntology.ai/api/ran/2310.15047.svg)](https://syntology.ai/paper/2310.15047)

Connect an agent — have() is free.