SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2603.26680 · 2026

AlpsBench: An LLM Personalization Benchmark for Real-Dialogue Memorization and Preference Alignment

Yang Zhang, Xiangnan He, Fuli Feng, Wenjie Wang, Xinyu Lin, Xiang Yu, Chengbing Wang, Hongxun Ding, Jianfei Xiao, Wuqiang Zheng, Kaining Liu

arXiv · PDF · Open in the Atlas

Code that ran

8 of 8 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 3 have no contract verdict; 5 failures of our instrument; 0 no recorded run), measured 2026-11-22; 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/2603.26680.json · badge: /api/ran/2603.26680.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
ThisIsCosine/AlpsBench community
Papers with Code archive
8 ran · 0 no recorded run
FunctionStatusWhere it lives
_candidate_alps_data_roots Ran
ran on a synthesized input
ThisIsCosine/AlpsBench/src/benchmark/data.py
code served (permissive licence) · get_code("dc698434416c0f93")
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, ndim 1, 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 os
from pathlib import Path
import tempfile

# The function takes no parameters, so ARGS should be an empty dict
ARGS = {}
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 list, ndim 1, dtype family float — a list 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
_task1_benchmark_id Ran
ran; our contract draft was wrong, not the code
ThisIsCosine/AlpsBench/src/benchmark/data.py
code served (permissive licence) · get_code("19ef552fcf2e2516")
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 str. 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

torch.manual_seed(0)

ARGS = {
    "session_id": "session_001",
    "canonical_id": "canonical_abc123"
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
_task1_dialogue_from_sessions Ran
ran; our contract draft was wrong, not the code
ThisIsCosine/AlpsBench/src/benchmark/data.py
code served (permissive licence) · get_code("9a5f9d50b5f4b241")
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, ndim 1, 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.

# Setup code for _task1_dialogue_from_sessions

ARGS = {
    'sessions': [
        {
            'turns': [
                {'role': 'user', 'text': 'Hello, how are you?'},
                {'role': 'assistant', 'text': 'I am doing well, thank you!'},
                {'role': 'user', 'text': 'What is your name?'},
                {'role': 'assistant', 'text': 'I am Claude.'},
            ]
        },
        {
            'turns': [
                {'role': 'user', 'text': 'This is ignored'},
            ]
        }
    ]
}
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 list, ndim 1, dtype family float — a list 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
_task1_manifest_path Ran
ran; our contract draft was wrong, not the code
ThisIsCosine/AlpsBench/src/benchmark/data.py
code served (permissive licence) · get_code("e8b640ff11be13c5")
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 PosixPath. 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.

from pathlib import Path

ARGS = {
    'alps_data_root': Path('/tmp/alps_data')
}
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 PosixPath — a PosixPath output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a PosixPath 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
_task1_output_split_dir Ran
ran; our contract draft was wrong, not the code
ThisIsCosine/AlpsBench/src/benchmark/data.py
code served (permissive licence) · get_code("f9d5b4e2e75131aa")
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 PosixPath. 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.

from pathlib import Path

ARGS = {
    'alps_data_root': Path('/tmp/alps_data')
}
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 PosixPath — a PosixPath output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a PosixPath 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
load_task1_authoritative_pairs Ran
ran on a synthesized input
ThisIsCosine/AlpsBench/src/benchmark/data.py
code served (permissive licence) · get_code("562d0b743158bbfd")
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) and returned list, ndim 2, 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 os
from pathlib import Path
from typing import Any
import json
import tempfile
import shutil

# Create a temporary directory structure that mimics the Alps data layout
temp_dir = tempfile.mkdtemp()
alps_data_root = Path(temp_dir) / "alps_data"
alps_data_root.mkdir(parents=True, exist_ok=True)

# Create the manifest directory and file
manifest_dir = alps_data_root / 'manual_check_full_usable_dataset' / 'benchmark_construction_check_data' / 'task1'
manifest_dir.mkdir(parents=True, exist_ok=True)

# Create sample manifest data
manifest_data = [
    {
        'session_id': 'sess_001',
        'canonical_id': 'canon_001',
        'stratum': 'TIER1'
    },
    {
        'session_id': 'sess_002',
        'canonical_id': 'canon_002',
        'stratum': 'TIER2'
    }
]

manifest_path = manifest_dir / 'manifest.json'
with manifest_path.open('w', encoding='utf-8') as f:
    json.dump(manifest_data, f)

# Create the output_split directory with split files
output_split_dir = alps_data_root / 'output_split' / 'output_split'
output_split_dir.mkdir(parents=True, exist_ok=True)

# Create sample split data
split_data = {
    'data': [
        {
            'user_id': 'user_123',
            'line_index': 0,
            'sessions': [
                {
                    'session_id': 'sess_001',
                    'turns': [
                        {'role': 'user', 'text': 'Hello'},
                        {'role': 'assistant', 'text': 'Hi there'}
                    ]
                }
            ],
            'match': 'exact',
            'memory_items': [{'id': 'mem_1', 'content': 'memory content 1'}],
            'memory_stage1_candidates': [{'id': 'cand_1', 'score': 0.9}],
            'selected_memory_id': 'mem_1'
        },
        {
            'user_id': 'user_456',
            'line_index': 1,
            'sessions': [
                {
                    'session_id': 'sess_002',
                    'turns': [
                        {'role': 'user', 'text': 'How are you?'},
                        {'role': 'assistant', 'text': 'I am doing well'}
                    ]
                }
            ],
            'match': 'partial',
            'memory_items': [{'id': 'mem_2', 'content': 'memory content 2'}],
            'memory_stage1_candidates': [{'id': 'cand_2', 'score': 0.8}],
            'selected_memory_id': 'mem_2'
        }
    ]
}

split_file = output_split_dir / 'split_part_0.json'
with split_file.open('w', encoding='utf-8') as f:
    json.dump(split_data, f)

# Set the environment variable to point to our test data
os.environ['ALPS_DATA_ROOT'] = str(alps_data_root)

ARGS = {}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 list, ndim 2, dtype family float — a list 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
read_json Ran
ran; our contract draft was wrong, not the code
ThisIsCosine/AlpsBench/src/benchmark/data.py
code served (permissive licence) · get_code("fa52501efd960421")
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) 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 tempfile
import json
from pathlib import Path

# Create a temporary directory and file for testing
temp_dir = tempfile.TemporaryDirectory()
temp_path = Path(temp_dir.name) / "test.json"

# Write a valid JSON file
test_data = {
    "key1": "value1",
    "key2": [1, 2, 3],
    "key3": {"nested": "object"}
}

with temp_path.open('w', encoding='utf-8') as f:
    json.dump(test_data, f)

ARGS = {
    "path": temp_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 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
resolve_alps_data_root Ran
ran on a synthesized input
ThisIsCosine/AlpsBench/src/benchmark/data.py
code served (permissive licence) · get_code("518ee028930b5a35")
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 PosixPath. 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 os
from pathlib import Path
import tempfile

# Create a temporary directory to act as ALPS_DATA_ROOT
temp_dir = tempfile.mkdtemp()
os.environ['ALPS_DATA_ROOT'] = temp_dir

# The function resolve_alps_data_root takes no parameters
ARGS = {}
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 PosixPath — a PosixPath output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a PosixPath 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 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 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 8 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

As Large Language Models (LLMs) evolve into lifelong AI assistants, LLM personalization has become a critical frontier. However, progress is currently bottlenecked by the absence of a gold-standard evaluation benchmark. Existing benchmarks either overlook personalized information management that is critical for personalization or rely heavily on synthetic dialogues, which exhibit an inherent distribution gap from real-world dialogue. To bridge this gap, we introduce AlpsBench, An LLM PerSonalization benchmark derived from real-world human-LLM dialogues. AlpsBench comprises 2,500 long-term interaction sequences curated from Wild-Chat, paired with human-verified structured memories that encapsulate both explicit and implicit personalization signals. We define four pivotal tasks-personalized information extraction, updating, retrieval, and utilization-and establish protocols to evaluate the entire lifecycle of memory management. Our benchmarking of frontier LLMs and memory-centric systems reveals that: (i) models struggle to reliably extract latent user traits; (ii) memory updating faces a performance ceiling even in the strongest models; (iii) retrieval accuracy declines sharply in the presence of large distractor pools; and (iv) while explicit memory mechanisms improve recall, they do not inherently guarantee more preference-aligned or emotionally resonant responses. AlpsBench aims to provide a comprehensive framework to accelerate research toward truly personalized AI assistants. • Information systems → Personalization.

For agents

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

get_harvested_code_for_paper("2603.26680")
get_code_for_paper("2603.26680")
have("2603.26680")

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

curl https://syntology.ai/api/ran/2603.26680.json

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

[![Syntology run record](https://syntology.ai/api/ran/2603.26680.svg)](https://syntology.ai/paper/2603.26680)

Connect an agent — have() is free.