SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2306.16680 · 2023

Exploring the Representation Power of SPLADE Models

arXiv · PDF · Open in the Atlas

Code that ran

5 of 5 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; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 4 failures 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/2306.16680.json · badge: /api/ran/2306.16680.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
ielab/understanding-splade canonical (classifier label)
classifier basis: census
5 ran · 0 no recorded run
FunctionStatusWhere it lives
combine_faiss_results Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ielab/understanding-splade/tevatron/src/tevatron/faiss_retriever/reducer.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from gangiswag/llm-reranker, is pointer only (licence: NONE) · get_code("9adb9c459cc4c8f0")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from gangiswag/llm-reranker, and GitHub detected no licence in that repository 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_dense_rep Ran
ran, honoured the contract we drafted
ielab/understanding-splade/tevatron/src/tevatron/trainer.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from gangiswag/llm-reranker, is pointer only (licence: NONE) · get_code("fd973ee45608bfd5")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from gangiswag/llm-reranker, and GitHub detected no licence in that repository 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.

pickle_load Ran
ran; our contract draft was wrong, not the code
ielab/understanding-splade/tevatron/src/tevatron/faiss_retriever/__main__.py
code served (permissive licence) · get_code("d5944ba894d22f55")
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 pickle
import numpy as np
import tempfile
import os

# Create a temporary pickle file with test data
temp_dir = tempfile.mkdtemp()
test_path = os.path.join(temp_dir, "test.pkl")

# Create sample data
reps = [[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0]]
lookup = {"key1": 0, "key2": 1}

# Write the pickle file
with open(test_path, 'wb') as f:
    pickle.dump((reps, lookup), f)

ARGS = {
    "path": test_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — 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
search_queries Ran
ran; our contract draft was wrong, not the code
ielab/understanding-splade/tevatron/src/tevatron/faiss_retriever/__main__.py
code served (permissive licence) · get_code("01c4c73be2b9b863")
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, 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

# Small stand-in retriever class
class MockRetriever:
    def __init__(self):
        pass
    
    def batch_search(self, q_reps, depth, batch_size, quiet):
        # q_reps shape: (num_queries, embedding_dim)
        num_queries = q_reps.shape[0]
        # Return scores and indices
        all_scores = np.random.rand(num_queries, depth).astype(np.float32)
        all_indices = np.random.randint(0, 100, size=(num_queries, depth))
        return all_scores, all_indices
    
    def search(self, q_reps, depth):
        num_queries = q_reps.shape[0]
        all_scores = np.random.rand(num_queries, depth).astype(np.float32)
        all_indices = np.random.randint(0, 100, size=(num_queries, depth))
        return all_scores, all_indices

# Small args config class
class Args:
    def __init__(self, batch_size=32, depth=10, quiet=False):
        self.batch_size = batch_size
        self.depth = depth
        self.quiet = quiet

# Create retriever
retriever = MockRetriever()

# Create query representations (num_queries=2, embedding_dim=4)
q_reps = np.random.rand(2, 4).astype(np.float32)

# Create passage lookup dictionary mapping passage indices to passage IDs
p_lookup = {i: f"passage_{i}" for i in range(100)}

# Create args
args = Args(batch_size=32, depth=10, quiet=False)

ARGS = {
    "retriever": retriever,
    "q_reps": q_reps,
    "p_lookup": p_lookup,
    "args": args
}
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 tuple, ndim 3, dtype family float — a tuple 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
split_dense_inputs Ran
ran; our contract draft was wrong, not the code
ielab/understanding-splade/tevatron/src/tevatron/trainer.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from gangiswag/llm-reranker, is pointer only (licence: NONE) · get_code("a68ca4e05b568068")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from gangiswag/llm-reranker, and GitHub detected no licence in that repository 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.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 5 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 5 later) and 0 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
get_dense_rep 4 other papers
2 dated earlier, 2 later
ijcai2022_0275 (2022 (year only))
Towards Robust Dense Retrieval via Local Ranking Alignment
2304.11370 (2023-04)
SAILER: Structure-aware Pre-trained Language Model for Legal Case Retrieval
2024.findings-emnlp.625 (2024 (year only))
STARD: A Chinese Statute Retrieval Dataset Derived from Real-life Queries by Non-professionals
and 1 more not named here
split_dense_inputs 4 other papers
2 dated earlier, 2 later
ijcai2022_0275 (2022 (year only))
Towards Robust Dense Retrieval via Local Ranking Alignment
2304.11370 (2023-04)
SAILER: Structure-aware Pre-trained Language Model for Legal Case Retrieval
2024.findings-emnlp.625 (2024 (year only))
STARD: A Chinese Statute Retrieval Dataset Derived from Real-life Queries by Non-professionals
and 1 more not named here
search_queries 2 other papers
2 later
2404.18424 (2024-04)
PromptReps: Prompting Large Language Models to Generate Dense and Sparse Representations for Zero-Shot Document Retrieval
2410.13230 (2024-10)
Starbucks-v2: Improved Training for 2D Matryoshka Embeddings
combine_faiss_results 1 other paper
1 later
2406.15657 (2024-06)
FIRST: Faster Improved Listwise Reranking with Single Token Decoding
pickle_load 1 other paper
1 later
2404.18424 (2024-04)
PromptReps: Prompting Large Language Models to Generate Dense and Sparse Representations for Zero-Shot Document Retrieval

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

The SPLADE (SParse Lexical AnD Expansion) model is a highly effective approach to learned sparse retrieval, where documents are represented by term impact scores derived from large language models. During training, SPLADE applies regularization to ensure postings lists are kept sparse -- with the aim of mimicking the properties of natural term distributions -- allowing efficient and effective lexical matching and ranking. However, we hypothesize that SPLADE may encode additional signals into common postings lists to further improve effectiveness. To explore this idea, we perform a number of empirical analyses where we re-train SPLADE with different, controlled vocabularies and measure how effective it is at ranking passages. Our findings suggest that SPLADE can effectively encode useful ranking signals in documents even when the vocabulary is constrained to terms that are not traditionally useful for ranking, such as stopwords or even random words.

For agents

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

get_harvested_code_for_paper("2306.16680")
get_code_for_paper("2306.16680")
have("2306.16680")

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

curl https://syntology.ai/api/ran/2306.16680.json

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

[![Syntology run record](https://syntology.ai/api/ran/2306.16680.svg)](https://syntology.ai/paper/2306.16680)

Connect an agent — have() is free.