SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2502.08056 · 2025

Cognify: Supercharging Gen-AI Workflows With Hierarchical Autotuning

arXiv · PDF · Open in the Atlas

Code that ran

4 of 4 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 4 have no contract verdict; 0 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/2502.08056.json · badge: /api/ran/2502.08056.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
GenseeAI/cognify canonical (classifier label)
classifier basis not recorded on the link
4 ran · 0 no recorded run
FunctionStatusWhere it lives
get_function_kwargs Ran
ran on a synthesized input
GenseeAI/cognify/cognify/graph/utils.py
code served (permissive licence) · get_code("f6fcb4b0ad7fd709")
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 inspect
from typing import Callable

def dummy_function(x: int, y: str = "default", z: float = 3.14):
    """A dummy function for testing get_function_kwargs."""
    pass

def another_function(a, b=10, c=None):
    """Another test function."""
    pass

def function_with_varargs(x, *args, y=5, **kwargs):
    """This should raise an error."""
    pass

# Test the function itself
ARGS = {
    "func": dummy_function
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — 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
hint_possible_destinations Ran
ran on a synthesized input
GenseeAI/cognify/cognify/graph/base.py
code served (permissive licence) · get_code("41c7da7713bd43f5")
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 function. 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

ARGS = {
    'dests': ['destination1', 'destination2', 'destination3']
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
strongly_connected_components Ran
ran on a synthesized input
GenseeAI/cognify/cognify/graph/utils.py
code served (permissive licence) · get_code("66ce939ff458dbaf")
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. 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 numpy as np

# Analyze the function: strongly_connected_components takes a graph parameter
# The graph is used as a dictionary where:
# - Keys are nodes
# - Values (graph[node]) are iterables of successor nodes
# This implements Tarjan's algorithm for finding strongly connected components

# Create a simple graph as a dictionary
# Nodes can be any hashable type - let's use integers
# Example: a graph with some cycles to demonstrate SCCs

graph = {
    0: [1],
    1: [2],
    2: [0, 3],
    3: [4],
    4: [3, 5],
    5: []
}

# This graph has two strongly connected components:
# - {0, 1, 2} (cycle)
# - {3, 4} (cycle)
# - {5} (singleton)

ARGS = {
    "graph": graph
}
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 — a list 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
subgraph Ran
ran on a synthesized input
GenseeAI/cognify/cognify/graph/utils.py
code served (permissive licence) · get_code("140b752c29fd791f")
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 dict. 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 numpy as np

# The function expects G to be a dict-like object where G[v] returns a set-like object
# that supports the & (intersection) operator with vertices
# vertices should be an iterable of vertex keys

# Create a simple graph representation
# G[v] will be a set of adjacent vertices for each vertex v
G = {
    'a': {'b', 'c'},
    'b': {'a', 'c', 'd'},
    'c': {'a', 'b'},
    'd': {'b'}
}

# vertices is the subset of vertices we want the subgraph for
vertices = {'a', 'b', 'c'}

ARGS = {
    'G': G,
    'vertices': vertices
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 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 4 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

Today's gen-AI workflows that involve multiple ML model calls, tool/API calls, data retrieval, or generic code execution are often tuned manually in an ad-hoc way that is both time-consuming and error-prone. In this paper, we propose a systematic approach for automatically tuning gen-AI workflows. Our key insight is that gen-AI workflows can benefit from structure, operator, and prompt changes, but unique properties of gen-AI workflows require new optimization techniques. We propose AdaSeek, an adaptive hierarchical search algorithm for autotuning gen-AI workflows. AdaSeek organizes workflow tuning methods into different layers based on the user-specified total search budget and distributes the budget across different layers based on the complexity of each layer. During its hierarchical search, AdaSeek redistributes the search budget from less useful to more promising tuning configurations based on workflow-level evaluation results. We implement AdaSeek in a workflow autotuning framework called Cognify and evaluate Cognify using six types of workflows such as RAG-based QA and text-to-SQL transformation. Overall, Cognify improves these workflows' generation quality by up to 2.8x, reduces execution monetary cost by up to 10x, and reduces end-to-end latency by 2.7x.

For agents

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

get_harvested_code_for_paper("2502.08056")
get_code_for_paper("2502.08056")
have("2502.08056")

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

curl https://syntology.ai/api/ran/2502.08056.json

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

[![Syntology run record](https://syntology.ai/api/ran/2502.08056.svg)](https://syntology.ai/paper/2502.08056)

Connect an agent — have() is free.