SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2106.05304 · ICML · 2021

Revisiting Point Cloud Shape Classification with a Simple and Effective Baseline

Jia Deng, Ankit Goyal, Hei Law, Bowei Liu, Alejandro Newell

arXiv · PDF · Open in the Atlas

Code that ran

4 of 12 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (3 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; 1 failure of our instrument; 8 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/2106.05304.json · badge: /api/ran/2106.05304.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
princeton-vl/SimpleView canonical (classifier label)
classifier basis: classifier_2025
1 ran · 8 no recorded run
axeber01/point-tnt community
Papers with Code archive
3 ran · 0 no recorded run
FunctionStatusWhere it lives
Attention Ran
ran on a synthesized input · constructed an object; no result was computed
axeber01/point-tnt/model.py
code served (permissive licence) · get_code("15cfb9f4479acd04")
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 an instance of its own class, Attention: the run constructed the object and did not call it, so no result was computed. 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

torch.manual_seed(0)

# The Attention module expects:
# - dim: input embedding dimension
# - heads: number of attention heads (default 3)
# - dim_head: dimension per head (default 64)
# - dropout: dropout rate (default 0.0)

ARGS = {
    "dim": 4,
    "heads": 2,
    "dim_head": 8,
    "dropout": 0.1,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
einops, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Attention — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Attention 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
Baseline Ran
ran on a synthesized input · constructed an object; no result was computed
axeber01/point-tnt/model.py
code served (permissive licence) · get_code("fe12c37bc8cebc08")
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 an instance of its own class, Baseline: the run constructed the object and did not call it, so no result was computed. 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
from torch import nn
from einops.layers.torch import Rearrange

torch.manual_seed(0)

class Attention(nn.Module):

    def __init__(self, *, dim, heads=3, dim_head=64, dropout=0.0):
        super().__init__()
        inner_dim = heads * dim_head
        self.heads = heads
        self.scale = dim_head ** (-0.5)
        self.to_qkv = nn.Linear(dim, inner_dim * 3, bias=False)
        self.to_out = nn.Sequential(nn.Linear(inner_dim, dim), nn.Dropout(dropout))

    def forward(self, x):
        from einops import rearrange, einsum as einops_einsum
        b, n, d, h = (*x.shape, self.heads)
        q, k, v = self.to_qkv(x).chunk(3, dim=-1)
        q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v))
        sim = torch.einsum('b i d, b j d -> b i j', q, k) * self.scale
        attn = sim.softmax(dim=-1)
        out = torch.einsum('b i j, b j d -> b i d', attn, v)
        out = rearrange(out, '(b h) n d -> b n (h d)', h=h)
        return self.to_out(out)

class FeedForward(nn.Module):

    def __init__(self, dim, mult=4, dropout=0.0):
        super().__init__()
        self.net = nn.Sequential(nn.Linear(dim, dim * mult), nn.GELU(), nn.Dropout(dropout), nn.Linear(dim * mult, dim))

    def forward(self, x):
        return self.net(x)

class PreNorm(nn.Module):

    def __init__(self, dim, fn):
        super().__init__()
        self.norm = nn.LayerNorm(dim)
        self.fn = fn

    def forward(self, x, **kwargs):
        return self.fn(self.norm(x), **kwargs)

class ArgsConfig:
    def __init__(self):
        self.global_attention = True
        self.patch_dim = 8
        self.depth = 2
        self.heads = 2
        self.dim_head = 4
        self.attn_dropout = 0.0
        self.ff_dropout = 0.0
        self.emb_dims = 16

class Baseline(nn.Module):
    """
    Baseline Transformer model that applies self-attention globally on all points
    """

    def __init__(self, args, num_classes, channels=3):
        super(Baseline, self).__init__()
        self.global_attention = args.global_attention
        self.to_anchor = nn.Sequential(Rearrange('b c n -> b n c'), nn.Linear(channels, args.patch_dim))
        layers = nn.ModuleList([])
        for _ in range(args.depth):
            layers.append(nn.ModuleList([PreNorm(args.patch_dim, Attention(dim=args.patch_dim, heads=args.heads, dim_head=args.dim_head, dropout=args.attn_dropout)) if self.global_attention else nn.Identity(), PreNorm(args.patch_dim, FeedForward(dim=args.patch_dim, dropout=args.ff_dropout))]))
        self.layers = layers
        self.final_conv = nn.Sequential(nn.LayerNorm(args.patch_dim * args.depth), nn.Linear(args.patch_dim * args.depth, args.emb_dims), nn.GELU())
        self.mlp_head = nn.Sequential(nn.Linear(args.emb_dims * 2, 512), nn.BatchNorm1d(512), nn.GELU(), nn.Dropout(0.5), nn.Linear(512, 256), nn.BatchNorm1d(256), nn.GELU(), nn.Dropout(0.5), nn.Linear(256, num_classes))

    def forward(self, x):
        batch_size, _, _ = x.shape
        points = self.to_anchor(x)
        ylist = []
        for patch_attn, patch_ff in self.layers:
            points = patch_attn(points) + points
            points = patch_ff(points) + points
            ylist.append(points)
        y = torch.cat(ylist, dim=-1)
        y = self.final_conv(y)
        y1 = y.max(dim=1, keepdim=False)[0]
        y2 = y.mean(dim=1, keepdim=False)
        y = torch.cat((y1, y2), dim=1)
        return self.mlp_head(y)

args = ArgsConfig()
num_classes = 10
channels = 3

ARGS = {
    "args": args,
    "num_classes": num_classes,
    "channels": channels
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
einops, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Baseline — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Baseline 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
FeedForward Ran
ran on a synthesized input · constructed an object; no result was computed
axeber01/point-tnt/model.py
code served (permissive licence) · get_code("bea49fdbe8f8578a")
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 an instance of its own class, FeedForward: the run constructed the object and did not call it, so no result was computed. 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

torch.manual_seed(0)

# FeedForward expects:
# - dim: int (feature dimension)
# - mult: int (multiplier for hidden layer, default 4)
# - dropout: float (dropout rate, default 0.0)

ARGS = {
    'dim': 4,
    'mult': 4,
    'dropout': 0.0,
}
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 FeedForward — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a FeedForward 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
fixed_padding Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
princeton-vl/SimpleView/ScanObjectNN/SimpleView/multi_model.py
code served (permissive licence) · get_code("ca82893aff5e0d2e")
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 EagerTensor, ndim 4, dtype family float, shape (2, 6, 7, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape 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 tensorflow as tf
import numpy as np

# Create a small tensor for testing
# Shape: [batch=2, channels=3, height=4, width=5] for channels_first
# or [batch=2, height=4, width=5, channels=3] for channels_last

batch_size = 2
height = 4
width = 5
channels = 3
kernel_size = 3

# Test with channels_first format
inputs_channels_first = tf.constant(
    np.random.randn(batch_size, channels, height, width),
    dtype=tf.float32
)

# Test with channels_last format
inputs_channels_last = tf.constant(
    np.random.randn(batch_size, height, width, channels),
    dtype=tf.float32
)

ARGS = {
    "inputs": inputs_channels_last,
    "kernel_size": kernel_size,
    "data_format": "channels_last"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, ndim 4, dtype family float, shape (2, 6, 7, 3) — shape (2, 6, 7, 3) as a later run recorded it
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
Model Not yet run
unverified (harvested, no recorded run)
princeton-vl/SimpleView/ScanObjectNN/SimpleView/multi_model.py
code served (permissive licence) · get_code("0aada2a0a966a06e")
How it ran

no invocation recorded.

_bottleneck_block_v1 Not yet run
unverified (harvested, no recorded run)
princeton-vl/SimpleView/ScanObjectNN/SimpleView/multi_model.py
code served (permissive licence) · get_code("b941b66d21901b83")
How it ran

no invocation recorded.

_bottleneck_block_v2 Not yet run
unverified (harvested, no recorded run)
princeton-vl/SimpleView/ScanObjectNN/SimpleView/multi_model.py
code served (permissive licence) · get_code("bd2e7dfc294b27b6")
How it ran

no invocation recorded.

_building_block_v1 Not yet run
unverified (harvested, no recorded run)
princeton-vl/SimpleView/ScanObjectNN/SimpleView/multi_model.py
code served (permissive licence) · get_code("11702804c488ed55")
How it ran

no invocation recorded.

_building_block_v2 Not yet run
unverified (harvested, no recorded run)
princeton-vl/SimpleView/ScanObjectNN/SimpleView/multi_model.py
code served (permissive licence) · get_code("d4516530837fed03")
How it ran

no invocation recorded.

batch_norm Not yet run
unverified (harvested, no recorded run)
princeton-vl/SimpleView/ScanObjectNN/SimpleView/multi_model.py
code served (permissive licence) · get_code("6cccb6ae78594ebb")
How it ran

no invocation recorded.

block_layer Not yet run
unverified (harvested, no recorded run)
princeton-vl/SimpleView/ScanObjectNN/SimpleView/multi_model.py
code served (permissive licence) · get_code("5425aca7fe65bb0d")
How it ran

no invocation recorded.

conv2d_fixed_padding Not yet run
unverified (harvested, no recorded run)
princeton-vl/SimpleView/ScanObjectNN/SimpleView/multi_model.py
code served (permissive licence) · get_code("a5cb548e4b581359")
How it ran

no invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 12 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 1 later) and 10 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
fixed_padding
brick of this paper's assembly: helper
4 other papers
3 dated earlier, 1 same year, order not determinable
1811.06964 (2018-11)
Grasp2Vec: Learning Object Representations from Self-Supervised Grasping
1812.00442 (2018-12)
Deep Cosine Metric Learning for Person Re-Identification
1906.01444 (2019-06)
Heterogeneous Gaussian Mechanism: Preserving Differential Privacy in Deep Learning with Provable Robustness
and 1 more not named here
FeedForward 2 other papers
1 dated earlier, 1 later · 2 only as an assembly brick
2103.00112 (2021-03)
Transformer in Transformer
2203.08913 (2022-03)
Memorizing Transformers

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

Processing point cloud data is an important component of many real-world systems. As such, a wide variety of point-based approaches have been proposed, reporting steady benchmark improvements over time. We study the key ingredients of this progress and uncover two critical results. First, we find that auxiliary factors like different evaluation schemes, data augmentation strategies, and loss functions, which are independent of the model architecture, make a large difference in performance. The differences are large enough that they obscure the effect of architecture. When these factors are controlled for, Point-Net++, a relatively older network, performs competitively with recent methods. Second, a very simple projection-based method, which we refer to as SimpleView, performs surprisingly well. It achieves on par or better results than sophisticated state-of-the-art methods on ModelNet40 while being half the size of PointNet++. It also outperforms state-of-the-art methods on ScanOb-jectNN, a real-world point cloud benchmark, and demonstrates better cross-dataset generalization. Code is available at https://github.com/ princeton-vl/SimpleView.

For agents

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

get_harvested_code_for_paper("2106.05304")
get_code_for_paper("2106.05304")
have("2106.05304")

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

curl https://syntology.ai/api/ran/2106.05304.json

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

[![Syntology run record](https://syntology.ai/api/ran/2106.05304.svg)](https://syntology.ai/paper/2106.05304)

Connect an agent — have() is free.