SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2410.05643 · ICLR · 2025

TRACE: Temporal Grounding Video LLM via Causal Event Modeling

Xi Chen, Yongxin Guo, Xiaoying Tang, Jingyu Liu, Mingda Li, Qingbin Liu

arXiv · PDF · Open in the Atlas

Code that ran

7 of 16 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, 1 violated it and 0 have no contract verdict; 6 failures of our instrument; 9 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/2410.05643.json · badge: /api/ran/2410.05643.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
gyxxyg/TRACE canonical (classifier label)
classifier basis: census
7 ran · 5 no recorded run
gyxxyg/trace canonical (classifier label)
classifier basis: classifier_2025
0 ran · 4 no recorded run
FunctionStatusWhere it lives
load_data Ran
ran; our contract draft was wrong, not the code
gyxxyg/TRACE/trace/eval/evaluate.py
code served (permissive licence) · get_code("741e5b0b3aa32a32")
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
import json
import tempfile

# Create a temporary directory and annotation file
temp_dir = tempfile.mkdtemp()

# Create a sample annotation file in COCO format
anno_data = {
    "annotations": [
        {"image_id": "xHr8X2Wpmno.mp4", "caption": "A person walking"},
        {"image_id": "anotherVideo.mp4", "caption": "A dog running"},
    ]
}

# Write the annotation file
split_name = "train"
file_path = os.path.join(temp_dir, f'{split_name}.caption_coco_format.json')
with open(file_path, 'w') as f:
    json.dump(anno_data, f)

# Create a simple args object with debug attribute
class Args:
    def __init__(self, debug=False):
        self.debug = debug

ARGS = {
    "args": Args(debug=False),
    "anno_path": temp_dir,
    "split": split_name
}
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
merge_scenes Ran
ran; our contract draft was wrong, not the code
gyxxyg/TRACE/trace/mm_utils.py
code served (permissive licence) · get_code("4b339b70f2a3e67e")
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 int; 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 numpy as np

class Cut:
    def __init__(self, frames):
        self._frames = frames
    
    def get_frames(self):
        return self._frames

# Create test data
cut_list = [
    Cut(10),
    Cut(25),
    Cut(45),
    Cut(60),
    Cut(80)
]

cut_scores = np.array([0.9, 0.3, 0.7, 0.2, 0.8])

scene_list = [
    (0, 10),
    (10, 25),
    (25, 45),
    (45, 60),
    (60, 80),
    (80, 100)
]

num_frames = 100

ARGS = {
    'cut_list': cut_list,
    'cut_scores': cut_scores,
    'scene_list': scene_list,
    'num_frames': num_frames,
    'max_scene_num': 4,
    'num_frame_per_scene': 8,
    'min_frames_per_scene': 30
}
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 list, ndim 2, dtype family int — 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
merge_scenes_not_exeed_max_scene_num Ran
ran; our contract draft was wrong, not the code
gyxxyg/TRACE/trace/mm_utils.py
code served (permissive licence) · get_code("1e366edf3b7d93ae")
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 numpy as np

class Cut:
    """Stand-in class for cut objects that have a get_frames() method"""
    def __init__(self, frames):
        self._frames = frames
    
    def get_frames(self):
        return self._frames

# Create test data
cut_list = [
    Cut([0, 100]),
    Cut([100, 200]),
    Cut([200, 300]),
    Cut([300, 400]),
]

cut_scores = np.array([0.5, 0.2, 0.8, 0.3], dtype=np.float32)

scene_list = [
    (0, 100),
    (100, 200),
    (200, 300),
    (300, 400),
    (400, 500),
]

max_scene_num = 3

ARGS = {
    "cut_list": cut_list,
    "cut_scores": cut_scores,
    "scene_list": scene_list,
    "max_scene_num": max_scene_num,
}
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
merge_seg_caps Ran
ran; our contract draft was wrong, not the code
gyxxyg/TRACE/trace/eval/evaluate.py
code served (permissive licence) · get_code("d650b26a59df9e29")
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

def merge_seg_caps(results):
    """merge mulple generated captions from a same video into paragraph."""
    merge_results = {}
    for jterm in results:
        vname = jterm['vname']
        cap = jterm['generated_cap']
        postfix = vname.split('.mp4')[-1]
        start_time, end_time = (float(postfix.split('_')[-2]), float(postfix.split('_')[-1]))
        vid = vname.split('.mp4')[0] + '.mp4'
        if vid not in merge_results:
            merge_results[vid] = []
        merge_results[vid].append({'timestamp': [start_time, end_time], 'caption': cap})
    return merge_results

# Construct valid arguments
results = [
    {
        'vname': 'video1.mp4_0.5_1.5',
        'generated_cap': 'A person is walking in the park'
    },
    {
        'vname': 'video1.mp4_1.5_2.5',
        'generated_cap': 'The person sits down on a bench'
    },
    {
        'vname': 'video2.mp4_0.0_1.0',
        'generated_cap': 'A dog is running'
    },
    {
        'vname': 'video2.mp4_1.0_2.0',
        'generated_cap': 'The dog catches a frisbee'
    }
]

ARGS = {
    'results': results
}
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
read_txt Ran
ran; our contract draft was wrong, not the code
gyxxyg/TRACE/trace/eval/evaluate.py
code served (permissive licence) · get_code("e49f0cf401af904b")
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 tempfile
import os

# Create a temporary text file with some content
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test.txt")

with open(temp_file, 'w') as f:
    f.write("Hello, World!\n")
    f.write("This is a second line\n")

# Define arguments for read_txt function
ARGS = {
    "path": temp_file
}
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 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
split_to_even_chunks Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
gyxxyg/TRACE/trace/trace_trainer.py
code served (permissive licence) · get_code("10893c4608c08075")
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 int; 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

torch.manual_seed(0)

# Analyzing the function:
# - indices: a list/sequence of indices that will be iterated over
# - lengths: a dict-like or list-like object that maps indices to their lengths (accessed via lengths[index])
# - num_chunks: an integer specifying how many chunks to split into

# Create sample data
indices = [0, 1, 2, 3, 4, 5]
lengths = {0: 10, 1: 15, 2: 8, 3: 12, 4: 9, 5: 11}
num_chunks = 3

ARGS = {
    "indices": indices,
    "lengths": lengths,
    "num_chunks": num_chunks,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — 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 int — 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
violates_moderation Ran
ran, violated the contract we drafted
gyxxyg/TRACE/trace/utils.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from Mi3-Lab/doScenes-VLM-Planning, is pointer only (licence: AGPL-3.0) · get_code("f9939a84b9a65279")
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 Mi3-Lab/doScenes-VLM-Planning, and AGPL-3.0, as GitHub detects it on that repository, is not in the permissive set Syntology serves inline. 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.

TraceMetaForCausalLM Not yet run
unverified (harvested, no recorded run)
gyxxyg/trace/trace/model/trace_arch.py
code served (permissive licence) · get_code("fe9f9161144b6f90")
How it ran

no invocation recorded.

build_logger Not yet run
unverified (harvested, no recorded run)
gyxxyg/TRACE/trace/utils.py
code served (permissive licence) · get_code("8dc94b9682ebebf9")
How it ran

no invocation recorded.

get_mm_adapter_state_maybe_zero_3 Not yet run
unverified (harvested, no recorded run)
gyxxyg/TRACE/trace/trace_trainer.py
code served (permissive licence) · get_code("bb35e3ac741bb2c9")
How it ran

no invocation recorded.

get_peft_state_maybe_zero_3 Not yet run
unverified (harvested, no recorded run)
gyxxyg/trace/trace/train_mt.py
code served (permissive licence) · get_code("fa1225dfac92bc0d")
How it ran

no invocation recorded.

get_peft_state_non_lora_maybe_zero_3 Not yet run
unverified (harvested, no recorded run)
gyxxyg/trace/trace/train_mt.py
code served (permissive licence) · get_code("1c53657305b66e9f")
How it ran

no invocation recorded.

load_pretrained_model Not yet run
unverified (harvested, no recorded run)
gyxxyg/TRACE/trace/model/builder.py
code served (permissive licence) · get_code("a2c9ee5dff5e2278")
How it ran

no invocation recorded.

maybe_zero_3 Not yet run
unverified (harvested, no recorded run)
gyxxyg/TRACE/trace/trace_trainer.py
code served (permissive licence) · get_code("735025744c1ab0cf")
How it ran

no invocation recorded.

maybe_zero_3 Not yet run
unverified (harvested, no recorded run)
gyxxyg/trace/trace/train_mt.py
code served (permissive licence) · get_code("616ffbdc154ed2d8")
How it ran

no invocation recorded.

pretty_print_semaphore Not yet run
unverified (harvested, no recorded run)
gyxxyg/TRACE/trace/utils.py
code served (permissive licence) · get_code("37899f22fb191b37")
How it ran

no invocation recorded.

Also present in other papers' code

Of 16 samples lifted from this paper's linked repositories, 16 carry a function identity and 0 do not. Of the 16 distinct functions with an identity, 12 are also present in other papers whose code we hold (12 with at least one paper dated earlier, 8 the same month, 8 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.

FunctionAlso present inEarliest-datedOthers named
violates_moderation 56 other papers
39 dated earlier, 2 the same month, 14 later, 1 same year, order not determinable
2106.06909 (2021-06)
GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10,000 Hours of Transcribed Audio
2308.03279 (2023-08)
UniversalNER: TARGETED DISTILLATION FROM LARGE LANGUAGE MODELS FOR OPEN NAMED ENTITY RECOGNITION
2308.13437 (2023-08)
Position-Enhanced Visual Instruction Tuning for Multimodal Large Language Models
and 53 more not named here
pretty_print_semaphore 54 other papers
36 dated earlier, 2 the same month, 15 later, 1 same year, order not determinable
2106.06909 (2021-06)
GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10,000 Hours of Transcribed Audio
2308.13437 (2023-08)
Position-Enhanced Visual Instruction Tuning for Multimodal Large Language Models
2309.07120 (2023-09)
Sight Beyond Text: Multi-Modal Training Enhances LLMs in Truthfulness and Ethics
and 51 more not named here
get_mm_adapter_state_maybe_zero_3 43 other papers
23 dated earlier, 5 the same month, 15 later
2304.05884 (2023-04)
Unicom: Universal and Compact Representation Learning for Image Retrieval
2304.08485 (2023-04)
Visual Instruction Tuning
2310.03744 (2023-10)
Improved Baselines with Visual Instruction Tuning
and 40 more not named here
maybe_zero_3 42 other papers
24 dated earlier, 4 the same month, 14 later
2304.05884 (2023-04)
Unicom: Universal and Compact Representation Learning for Image Retrieval
2304.08485 (2023-04)
Visual Instruction Tuning
2310.03744 (2023-10)
Improved Baselines with Visual Instruction Tuning
and 39 more not named here
split_to_even_chunks 40 other papers
20 dated earlier, 5 the same month, 15 later
2304.05884 (2023-04)
Unicom: Universal and Compact Representation Learning for Image Retrieval
2304.08485 (2023-04)
Visual Instruction Tuning
2310.03744 (2023-10)
Improved Baselines with Visual Instruction Tuning
and 37 more not named here
maybe_zero_3 37 other papers
19 dated earlier, 3 the same month, 15 later
2310.01779 (2023-10)
HallE-Control: Controlling Object Hallucination in Large Multimodal Models
2311.16839 (2023-11)
Beyond Hallucinations: Enhancing LVLMs through Hallucination-Aware Direct Preference Optimization
2312.02949 (2023-12)
LLaVA-Grounding: Grounded Visual Chat with Large Multimodal Models
and 34 more not named here
get_peft_state_maybe_zero_3 23 other papers
11 dated earlier, 3 the same month, 9 later
2310.01779 (2023-10)
HallE-Control: Controlling Object Hallucination in Large Multimodal Models
2401.02330 (2024-01)
LLaVA-Phi: Efficient Multi-Modal Assistant with Small Language Model
2402.12501 (2024-02)
Your Vision-Language Model Itself Is a Strong Filter: Towards High-Quality Instruction Tuning with Data Selection
and 20 more not named here
get_peft_state_non_lora_maybe_zero_3 23 other papers
12 dated earlier, 3 the same month, 8 later
2310.01779 (2023-10)
HallE-Control: Controlling Object Hallucination in Large Multimodal Models
2312.02949 (2023-12)
LLaVA-Grounding: Grounded Visual Chat with Large Multimodal Models
2401.02330 (2024-01)
LLaVA-Phi: Efficient Multi-Modal Assistant with Small Language Model
and 20 more not named here
load_data 2 other papers
2 dated earlier
2312.02051 (2023-12)
TimeChat: A Time-sensitive Multimodal Large Language Model for Long Video Understanding
2405.13382 (2024-05)
VTG-LLM: Integrating Timestamp Knowledge into Video LLMs for Enhanced Video Temporal Grounding
merge_seg_caps 2 other papers
2 dated earlier
2312.02051 (2023-12)
TimeChat: A Time-sensitive Multimodal Large Language Model for Long Video Understanding
2405.13382 (2024-05)
VTG-LLM: Integrating Timestamp Knowledge into Video LLMs for Enhanced Video Temporal Grounding
read_txt 2 other papers
2 dated earlier
2312.02051 (2023-12)
TimeChat: A Time-sensitive Multimodal Large Language Model for Long Video Understanding
2405.13382 (2024-05)
VTG-LLM: Integrating Timestamp Knowledge into Video LLMs for Enhanced Video Temporal Grounding
build_logger 1 other paper
1 dated earlier
2409.19603 (2024-09)
One Token to Seg Them All: Language Instructed Reasoning Segmentation in Videos

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

Video Temporal Grounding (VTG) is a crucial capability for video understanding models and plays a vital role in downstream tasks such as video browsing and editing. To effectively handle various tasks simultaneously and enable zero-shot prediction, there is a growing trend in employing video LLMs for VTG tasks. However, current video LLM-based methods rely exclusively on natural language generation, lacking the ability to model the clear structure inherent in videos, which restricts their effectiveness in tackling VTG tasks. To address this issue, this paper first formally introduces causal event modeling framework, which represents video LLM outputs as sequences of events, and predict the current event using previous events, video inputs, and textural instructions. Each event consists of three components: timestamps, salient scores, and textual captions. We then propose a novel task-interleaved video LLM called TRACE to effectively implement the causal event modeling framework in practice. The TRACEprocess visual frames, timestamps, salient scores, and text as distinct tasks, employing various encoders and decoding heads for each. Task tokens are arranged in an interleaved sequence according to the causal event modeling framework's formulation. Extensive experiments on various VTG tasks and datasets demonstrate the superior performance of TRACE compared to state-of-the-art video LLMs. Our model and code are avaliable at https://github.com/gyxxyg/TRACE.

For agents

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

get_harvested_code_for_paper("2410.05643")
get_code_for_paper("2410.05643")
have("2410.05643")

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

curl https://syntology.ai/api/ran/2410.05643.json

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

[![Syntology run record](https://syntology.ai/api/ran/2410.05643.svg)](https://syntology.ai/paper/2410.05643)

Connect an agent — have() is free.