SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.01869 · 2024

InferCept: Efficient Intercept Support for Augmented Large Language Model Inference

arXiv · PDF · Open in the Atlas

Code that ran

3 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 0 have no contract verdict; 3 failures of our instrument; 1 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/2402.01869.json · badge: /api/ran/2402.01869.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
wuklab/infercept canonical (classifier label)
classifier basis: census
3 ran · 1 no recorded run
FunctionStatusWhere it lives
sample_requests Ran
ran; our contract draft was wrong, not the code
wuklab/infercept/benchmarks/benchmark_api_serving.py
code served (permissive licence) · get_code("84a26133838eab27")
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 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 json
import tempfile
import torch
from transformers import PreTrainedTokenizerBase

torch.manual_seed(0)

class MockTokenizer(PreTrainedTokenizerBase):
    """Minimal tokenizer that returns fixed-length token sequences"""
    
    def __init__(self):
        self.model_max_length = 2048
    
    def __call__(self, texts, **kwargs):
        """Return mock input_ids for texts"""
        if isinstance(texts, str):
            texts = [texts]
        
        class TokenizerOutput:
            def __init__(self, input_ids):
                self.input_ids = input_ids
        
        # Create input_ids: each text gets a mock sequence based on its length
        input_ids = []
        for text in texts:
            # Generate deterministic but variable length sequences
            text_len = len(text)
            seq_len = 50 + (text_len % 100)  # Vary between 50-150 tokens
            seq = [100 + i for i in range(min(seq_len, 200))]  # Cap at 200
            input_ids.append(seq)
        
        return TokenizerOutput(input_ids)

# Create a temporary dataset file
dataset = [
    {
        "conversations": [
            {"value": "What is machine learning?"},
            {"value": "Machine learning is a subset of artificial intelligence that enables systems to learn from data."}
        ]
    },
    {
        "conversations": [
            {"value": "How does neural network work?"},
            {"value": "Neural networks are computing systems inspired by biological neural networks that constitute animal brains."}
        ]
    },
    {
        "conversations": [
            {"value": "What is deep learning?"},
            {"value": "Deep learning is part of a broader family of machine learning methods based on artificial neural networks."}
        ]
    },
    {
        "conversations": [
            {"value": "Explain transformers."},
            {"value": "Transformers are neural network architectures that use self-attention mechanisms to process sequences efficiently."}
        ]
    },
    {
        "conversations": [
            {"value": "Tell me about LLMs."},
            {"value": "Large language models are neural networks trained on vast amounts of text data to understand and generate human language."}
        ]
    }
]

with tempfile.NamedTemporaryFile(mode='w', suffix='.json', delete=False) as f:
    json.dump(dataset, f)
    dataset_path = f.name

tokenizer = MockTokenizer()
num_requests = 2

ARGS = {
    "dataset_path": dataset_path,
    "num_requests": num_requests,
    "tokenizer": tokenizer
}
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 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
sample_requests Ran
ran; our contract draft was wrong, not the code
wuklab/infercept/benchmarks/benchmark_serving.py
code served (permissive licence) · get_code("f4e08ad9325905dc")
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 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 json
import random
import tempfile
from transformers import PreTrainedTokenizerBase

class StandInTokenizer(PreTrainedTokenizerBase):
    def __init__(self):
        pass
    
    def __call__(self, texts):
        if isinstance(texts, str):
            texts = [texts]
        
        class TokenizerOutput:
            def __init__(self, input_ids):
                self.input_ids = input_ids
        
        result_ids = []
        for text in texts:
            word_count = len(text.split())
            token_count = max(5, word_count + random.randint(1, 3))
            result_ids.append(list(range(token_count)))
        
        return TokenizerOutput(result_ids)

with tempfile.NamedTemporaryFile(mode='w', suffix='.json', delete=False) as f:
    dataset = [
        {
            "conversations": [
                {"value": "What is machine learning?"},
                {"value": "Machine learning is a subset of artificial intelligence that enables systems to learn and improve from experience without being explicitly programmed."}
            ]
        },
        {
            "conversations": [
                {"value": "Tell me about neural networks"},
                {"value": "Neural networks are computing systems inspired by biological neural networks that constitute animal brains. They are fundamental to deep learning."}
            ]
        },
        {
            "conversations": [
                {"value": "How does backpropagation work?"},
                {"value": "Backpropagation is a method used in artificial neural networks to calculate a gradient that is needed in the calculation of the weights to be used in the network."}
            ]
        },
        {
            "conversations": [
                {"value": "What is"},
                {"value": "short"}
            ]
        }
    ]
    json.dump(dataset, f)
    dataset_path = f.name

tokenizer = StandInTokenizer()

ARGS = {
    "dataset_path": dataset_path,
    "num_requests": 2,
    "tokenizer": tokenizer
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
transformers — 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
sample_requests Ran
ran; our contract draft was wrong, not the code
wuklab/infercept/benchmarks/benchmark_throughput.py
code served (permissive licence) · get_code("ed11039cd60ea03f")
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 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 json
import random
import tempfile
from transformers import AutoTokenizer

# Create a minimal tokenizer stand-in that implements the required interface
class MinimalTokenizer:
    def __call__(self, texts):
        """Tokenize texts and return object with input_ids attribute."""
        class TokenizerOutput:
            def __init__(self, input_ids):
                self.input_ids = input_ids
        
        if isinstance(texts, str):
            texts = [texts]
        
        # Simple tokenization: split by spaces and convert to token ids
        input_ids = []
        for text in texts:
            tokens = text.split()
            # Create token ids of varying lengths based on text length
            token_ids = list(range(len(tokens))) if tokens else [0]
            input_ids.append(token_ids)
        
        return TokenizerOutput(input_ids)

# Create a temporary dataset file
dataset_content = [
    {
        "conversations": [
            {"value": "What is machine learning? " * 50},  # Long prompt to ensure > 4 tokens
            {"value": "Machine learning is a subset of artificial intelligence. " * 50}  # Long completion
        ]
    },
    {
        "conversations": [
            {"value": "How does deep learning work? " * 60},
            {"value": "Deep learning uses neural networks with multiple layers. " * 60}
        ]
    },
    {
        "conversations": [
            {"value": "Explain natural language processing. " * 55},
            {"value": "NLP is the field of AI that focuses on human language. " * 55}
        ]
    },
    {
        "conversations": [
            {"value": "What is a transformer model? " * 70},
            {"value": "Transformers are neural network architectures based on attention mechanisms. " * 70}
        ]
    },
]

# Write dataset to a temporary file
with tempfile.NamedTemporaryFile(mode='w', suffix='.json', delete=False) as f:
    json.dump(dataset_content, f)
    dataset_path = f.name

tokenizer = MinimalTokenizer()
num_requests = 2

ARGS = {
    "dataset_path": dataset_path,
    "num_requests": num_requests,
    "tokenizer": tokenizer,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
transformers — 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
run_hf Not yet run
unverified (harvested, no recorded run)
wuklab/infercept/benchmarks/benchmark_throughput.py
code served (permissive licence) · get_code("62f75e14e50e58ad")
How it ran

no invocation recorded.

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 2 distinct functions with an identity, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 1 the same month, 1 later) and 1 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
run_hf 3 other papers
1 the same month, 2 later
2402.01799 (2024-02)
Faster and Lighter LLMs: A Survey on Current Challenges and Way Forward
2405.16444 (2024-05)
CacheBlend: Fast Large Language Model Serving for RAG with Cached Knowledge Fusion
2410.00428 (2024-10)
LayerKV: Optimizing Large Language Model Serving with Layer-wise KV Cache Management

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

Large language models are increasingly integrated with external environments, tools, and agents like ChatGPT plugins to extend their capability beyond language-centric tasks. However, today's LLM inference systems are designed for standalone LLMs. They treat each external interaction as the end of LLM generation and form a new request when the interaction finishes, causing unnecessary recomputation of already computed contexts, which accounts for 37-40% of total model forwarding time. This paper presents InferCept, the first LLM inference framework targeting augmented LLMs and supporting the efficient interception of LLM generation. InferCept minimizes the GPU resource waste caused by LLM interceptions and dedicates saved memory for serving more requests. InferCept improves the overall serving throughput by 1.6x-2x and completes 2x more requests per second compared to the state-of-the-art LLM inference systems.

For agents

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

get_harvested_code_for_paper("2402.01869")
get_code_for_paper("2402.01869")
have("2402.01869")

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

curl https://syntology.ai/api/ran/2402.01869.json

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

[![Syntology run record](https://syntology.ai/api/ran/2402.01869.svg)](https://syntology.ai/paper/2402.01869)

Connect an agent — have() is free.