SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2106.01269 · ACL · 2021

More Identifiable yet Equally Performant Transformers for Text Classification

Eduard Hovy, Soujanya Poria, Rishabh Bhardwaj, Navonil Majumder

arXiv · PDF · Open in the Atlas

Code that ran

2 of 2 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 1 failure of our instrument; 0 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/2106.01269.json · badge: /api/ran/2106.01269.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
declare-lab/identifiable-transformers marked official in Papers with Code
Papers with Code archive
2 ran · 0 no recorded run
FunctionStatusWhere it lives
MultiheadAttention Ran
ran on a synthesized input · constructed an object; no result was computed
declare-lab/identifiable-transformers/model_identifiable.py
code served (permissive licence) · get_code("65062d0b9d191cb5")
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, MultiheadAttention: 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

torch.manual_seed(0)

# Parameters for MultiheadAttention.__init__
embed_dim = 8
num_heads = 2
kdim = 4
vdim = 4
concat_heads = True
dropout = 0.1
bias = True

ARGS = {
    'embed_dim': embed_dim,
    'num_heads': num_heads,
    'concat_heads': concat_heads,
    'dropout': dropout,
    'bias': bias,
    'kdim': kdim,
    'vdim': vdim,
}
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 MultiheadAttention — 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 MultiheadAttention 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
multi_head_attention_forward Ran
ran; our contract draft was wrong, not the code
declare-lab/identifiable-transformers/model_identifiable.py
code served (permissive licence) · get_code("650a5edfdd5ec496")
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 tuple. 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 functools import reduce

torch.manual_seed(0)

# Set up dimensions
tgt_len = 3  # sequence length
bsz = 2      # batch size
num_heads = 2
head_dim = 2
embed_dim = num_heads * head_dim  # 4

# Create projection weights
# q_proj_weight, k_proj_weight, v_proj_weight need shape (embed_dim, embed_dim)
q_proj_weight = torch.randn(embed_dim, embed_dim)
k_proj_weight = torch.randn(embed_dim, embed_dim)
v_proj_weight = torch.randn(embed_dim, embed_dim)

# Create input tensors
query = torch.randn(tgt_len, bsz, embed_dim)
key = torch.randn(tgt_len, bsz, embed_dim)
value = torch.randn(tgt_len, bsz, embed_dim)

# Create bias for input projection
# in_proj_bias should have size equal to sum of output dims: embed_dim + embed_dim + embed_dim
in_proj_bias = torch.randn(embed_dim + embed_dim + embed_dim)

# Create output projection weights and bias
out_proj_weight = torch.randn(embed_dim, embed_dim)
out_proj_bias = torch.randn(embed_dim)

# Create key padding mask (optional)
# Shape should be (bsz, tgt_len) with boolean or 0/1 values
key_padding_mask = torch.zeros(bsz, tgt_len, dtype=torch.bool)

ARGS = {
    "query": query,
    "key": key,
    "value": value,
    "head_dim": head_dim,
    "concat_head_output": True,
    "embed_dim_to_check": embed_dim,
    "num_heads": num_heads,
    "in_proj_bias": in_proj_bias,
    "dropout_p": 0.1,
    "out_proj_weight": out_proj_weight,
    "out_proj_bias": out_proj_bias,
    "training": True,
    "key_padding_mask": key_padding_mask,
    "need_weights": False,
    "q_proj_weight": q_proj_weight,
    "k_proj_weight": k_proj_weight,
    "v_proj_weight": v_proj_weight,
}
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 tuple — a tuple 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

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 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 2 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

Interpretability is an important aspect of the trustworthiness of a model's predictions. Transformer's predictions are widely explained by the attention weights, i.e., a probability distribution generated at its self-attention unit (head). Current empirical studies provide shreds of evidence that attention weights are not explanations by proving that they are not unique. A recent study showed theoretical justifications to this observation by proving the non-identifiability of attention weights. For a given input to a head and its output, if the attention weights generated in it are unique, we call the weights identifiable. In this work, we provide deeper theoretical analysis and empirical observations on the identifiability of attention weights. Ignored in the previous works, we find the attention weights are more identifiable than we currently perceive by uncovering the hidden role of the key vector. However, the weights are still prone to be non-unique attentions that make them unfit for interpretation. To tackle this issue, we provide a variant of the encoder layer that decouples the relationship between key and value vector and provides identifiable weights up to the desired length of the input. We prove the applicability of such variations by providing empirical justifications on varied text classification tasks. The implementations are available at https://github.com/declare-lab/ identifiable-transformers.

For agents

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

get_harvested_code_for_paper("2106.01269")
get_code_for_paper("2106.01269")
have("2106.01269")

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

curl https://syntology.ai/api/ran/2106.01269.json

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

[![Syntology run record](https://syntology.ai/api/ran/2106.01269.svg)](https://syntology.ai/paper/2106.01269)

Connect an agent — have() is free.