SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2210.07144 · ICML · 2023

Reprogramming Pretrained Language Models for Antibody Sequence Infilling

Pin-Yu Chen, Payel Das, Igor Melnyk, Vijil Chenthamarakshan, Amit Dhurandhar, Inkit Padhi, Devleena Das, English Encoder

arXiv · PDF · Open in the Atlas

Code that ran

5 of 7 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (4 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 5 have no contract verdict; 0 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/2210.07144.json · badge: /api/ran/2210.07144.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
ibm/reprogbert marked official in Papers with Code
Papers with Code archive
5 ran · 2 no recorded run
FunctionStatusWhere it lives
BertConfigProtein Ran
ran on a synthesized input · constructed an object; no result was computed
ibm/reprogbert/models/ReprogBert.py
code served (permissive licence) · get_code("a1a55d4e85621ec7")
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 an instance of its own class, BertConfigProtein: the run constructed the object and did not call it, so no result was computed. 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 torch.nn as nn
from transformers import BertConfig

torch.manual_seed(0)

ARGS = {
    "vocab_size_protein": 25,
    "vocab_size": 30522,
    "hidden_size": 768,
    "num_hidden_layers": 12,
    "num_attention_heads": 12,
    "intermediate_size": 3072,
    "hidden_act": "gelu",
    "hidden_dropout_prob": 0.1,
    "attention_probs_dropout_prob": 0.1,
    "max_position_embeddings": 512,
    "type_vocab_size": 2,
    "initializer_range": 0.02,
    "layer_norm_eps": 1e-12,
    "pad_token_id": 0,
    "position_embedding_type": "absolute",
    "use_cache": True,
    "classifier_dropout": None,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type BertConfigProtein — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a BertConfigProtein 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
ModifiedBertLMPredictionHead Ran
ran on a synthesized input · constructed an object; no result was computed
ibm/reprogbert/models/ReprogBert.py
code served (permissive licence) · get_code("6ca397d85cca0690")
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 an instance of its own class, ModifiedBertLMPredictionHead: the run constructed the object and did not call it, so no result was computed. 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
from torch import nn
from transformers.models.bert.modeling_bert import BertConfig

torch.manual_seed(0)

# Create a minimal config that satisfies the ModifiedBertLMPredictionHead requirements
config = BertConfig(
    vocab_size=8,
    vocab_size_protein=6,
    hidden_size=4,
    num_hidden_layers=2,
    num_attention_heads=2,
    intermediate_size=8,
)

ARGS = {
    "config": config,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ModifiedBertLMPredictionHead — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a ModifiedBertLMPredictionHead 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
ModifiedBertModel Ran
ran on a synthesized input · constructed an object; no result was computed
ibm/reprogbert/models/ReprogBert.py
code served (permissive licence) · get_code("67dd34f3a2ef4690")
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 an instance of its own class, ModifiedBertModel: the run constructed the object and did not call it, so no result was computed. 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 torch.nn as nn
from transformers.models.bert.modeling_bert import BertConfig

torch.manual_seed(0)

# Create a minimal BertConfig
config = BertConfig(
    vocab_size=100,
    vocab_size_protein=50,
    hidden_size=4,
    num_hidden_layers=2,
    num_attention_heads=2,
    intermediate_size=8,
    hidden_act="gelu",
    hidden_dropout_prob=0.1,
    attention_probs_dropout_prob=0.1,
    max_position_embeddings=512,
    type_vocab_size=2,
    initializer_range=0.02,
    layer_norm_eps=1e-12,
    pad_token_id=0,
    pad_token_id_prot=0,
    position_embedding_type="absolute",
)

ARGS = {
    "config": config,
    "add_pooling_layer": True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ModifiedBertModel — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a ModifiedBertModel 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
ModifiedBertOnlyMLMHead Ran
ran on a synthesized input
ibm/reprogbert/models/ReprogBert.py
code served (permissive licence) · get_code("448244116f48091e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, 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.

from transformers.models.bert.modeling_bert import BertLMPredictionHead, BertPreTrainedModel, BertEmbeddings, BertModel, BertOnlyMLMHead
from transformers import BertConfig
from torch import nn
import torch

class ModifiedBertLMPredictionHead(BertLMPredictionHead):

    def __init__(self, config):
        super().__init__(config)
        self.gamma = nn.Linear(config.vocab_size, config.vocab_size_protein, bias=False)
        self.gammabias = nn.Parameter(torch.zeros(config.vocab_size_protein))
        self.gamma.bias = self.gammabias

    def forward(self, hidden_states):
        hidden_states = super().forward(hidden_states)
        hidden_states = self.gamma(hidden_states)
        return hidden_states

class ModifiedBertOnlyMLMHead(BertOnlyMLMHead):

    def __init__(self, config):
        super().__init__(config)
        self.predictions = ModifiedBertLMPredictionHead(config)

    def forward(self, sequence_output):
        prediction_scores = self.predictions(sequence_output)
        return prediction_scores


config = BertConfig(
    vocab_size=32,
    hidden_size=8,
    num_hidden_layers=2,
    num_attention_heads=2,
    intermediate_size=16,
    hidden_act="gelu",
    hidden_dropout_prob=0.1,
    attention_probs_dropout_prob=0.1,
    max_position_embeddings=512,
    type_vocab_size=2,
    initializer_range=0.02,
    layer_norm_eps=1e-12,
)
config.vocab_size_protein = 24

INIT_ARGS = {
    "config": config
}

ARGS = {
    "sequence_output": torch.randn(2, 5, 8)
}

if __name__ == "__main__":
    model = ModifiedBertOnlyMLMHead(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Output shape: {output.shape}")
    print(f"Expected shape: (2, 5, 24)")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
ModifiedBertPreTrainedModel Ran
ran on a synthesized input · constructed an object; no result was computed
ibm/reprogbert/models/ReprogBert.py
code served (permissive licence) · get_code("6b4c0868aa3aca0e")
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 an instance of its own class, ModifiedBertPreTrainedModel: the run constructed the object and did not call it, so no result was computed. 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 torch.nn as nn
from transformers import BertConfig, BertPreTrainedModel

torch.manual_seed(0)

class BertConfigProtein(BertConfig):
    def __init__(self, vocab_size_protein=None, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self.vocab_size_protein = vocab_size_protein

class ModifiedBertPreTrainedModel(BertPreTrainedModel):
    config_class = BertConfigProtein

# Create a minimal BertConfigProtein instance
config = BertConfigProtein(
    vocab_size=30522,
    vocab_size_protein=25,
    hidden_size=4,
    num_hidden_layers=2,
    num_attention_heads=2,
    intermediate_size=8,
    hidden_act="gelu",
    hidden_dropout_prob=0.1,
    attention_probs_dropout_prob=0.1,
    max_position_embeddings=512,
    type_vocab_size=2,
    initializer_range=0.02,
    layer_norm_eps=1e-12,
)

ARGS = {
    "config": config,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ModifiedBertPreTrainedModel — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a ModifiedBertPreTrainedModel 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
BertForMaskedLMProt Not yet run
unverified (harvested, no recorded run)
ibm/reprogbert/models/ReprogBert.py
code served (permissive licence) · get_code("77372f4d71aefb84")
How it ran

no invocation recorded.

ModifiedBertEmbeddings Not yet run
unverified (harvested, no recorded run)
ibm/reprogbert/models/ReprogBert.py
code served (permissive licence) · get_code("7d1f150806b0e41d")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 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 7 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

Antibodies comprise the most versatile class of binding molecules, with numerous applications in biomedicine. Computational design of antibodies involves generating novel and diverse sequences, while maintaining structural consistency. Unique to antibodies, designing the complementaritydetermining region (CDR), which determines the antigen binding affinity and specificity, creates its own unique challenges. Recent deep learning models have shown impressive results, however the limited number of known antibody sequence/structure pairs frequently leads to degraded performance, particularly lacking diversity in the generated sequences. In our work we address this challenge by leveraging Model Reprogramming (MR), which repurposes pretrained models on a source language to adapt to the tasks that are in a different language and have scarce data -where it may be difficult to train a highperforming model from scratch or effectively finetune an existing pre-trained model on the specific task. Specifically, we introduce ReprogBert in which a pretrained English language model is repurposed for protein sequence infilling -thus considers cross-language adaptation using less data. Results on antibody design benchmarks show that our model on low-resourced antibody sequence dataset provides highly diverse CDR sequences, up to more than a two-fold increase of diversity over the baselines, without losing structural integrity and naturalness. The generated sequences also demonstrate enhanced antigen binding specificity and virus neutralization ability. Code is available at https://github. com/IBM/ReprogBERT

For agents

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

get_harvested_code_for_paper("2210.07144")
get_code_for_paper("2210.07144")
have("2210.07144")

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

curl https://syntology.ai/api/ran/2210.07144.json

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

[![Syntology run record](https://syntology.ai/api/ran/2210.07144.svg)](https://syntology.ai/paper/2210.07144)

Connect an agent — have() is free.