Mysterious behavior when grasping while controlling gripper poses using caterisan control

### Intro

Hi!

I am a Ph.d student at SDU, I use MuJoCo for my research on manipulation of deformable objects and reinforcement learning.

I am looking for any help, advice, insight or comments that might shine some light on my problem :)

Thanks in advance!

### My setup

I am running
- Ubuntu 24.04
- MuJoCo 3.3.7

### What's happening? What did you expect?

I have tried to use both the mocap and the actuator based control and get the results shown in the videos below

| Mocap Control | Actuator Control |
| ------------- | ------------- |
|  |  |

Here the mocap object behaves perfectly well, but there seems to be some odd phenomenon happening when I use the actuators. As you can see, when I rotate using the actuators on the revolute joints, it almost looks like the box is sliding forward? Like if the friction suddenly changes direction and moves the box ahead of the grippers fingers.

### Steps for reproduction

In order to reproduce this I have made attempted to only use the most default parameters and models. The script and model can be seen below

Simulation of both cases (change the `init` function when running)

```python
import time

import glfw
import mujoco
import mujoco as mj
import mujoco.viewer
import numpy as np
from robot_descriptions import robotiq_2f85_mj_description

_XML = """

"""

def add_act_freejoint(
spec: mj.MjSpec,
x_lim: tuple[float] = (-1, 1),
y_lim: tuple[float] = (-1, 1),
z_lim: tuple[float] = (-1, 1),
roll_lim: tuple[float] = (-np.pi, np.pi),
pitch_lim: tuple[float] = (-np.pi, np.pi),
yaw_lim: tuple[float] = (-np.pi, np.pi),
kp_pos: float = 1000,
kp_ori: float = 1000,
kv_pos: float = 1000,
kv_ori: float = 1000,
) -> mj.MjSpec:
base = spec.worldbody.first_body()
base.add_joint(name="x", axis=[1, 0, 0], type=mj.mjtJoint.mjJNT_SLIDE, range=x_lim)
base.add_joint(name="y", axis=[0, 1, 0], type=mj.mjtJoint.mjJNT_SLIDE, range=y_lim)
base.add_joint(name="z", axis=[0, 0, 1], type=mj.mjtJoint.mjJNT_SLIDE, range=z_lim)

base.add_joint(
name="roll", axis=[1, 0, 0], type=mj.mjtJoint.mjJNT_HINGE, range=roll_lim
)
base.add_joint(
name="pitch",
axis=[0, 1, 0],
type=mj.mjtJoint.mjJNT_HINGE,
range=pitch_lim,
)
base.add_joint(
name="yaw", axis=[0, 0, 1], type=mj.mjtJoint.mjJNT_HINGE, range=yaw_lim
)

spec.add_actuator(
name="x", target="x", trntype=mj.mjtTrn.mjTRN_JOINT, ctrlrange=x_lim
).set_to_position(kp=kp_pos, kv=kv_pos)

spec.add_actuator(
name="y", target="y", trntype=mj.mjtTrn.mjTRN_JOINT, ctrlrange=y_lim
).set_to_position(kp=kp_pos, kv=kv_pos)

spec.add_actuator(
name="z", target="z", trntype=mj.mjtTrn.mjTRN_JOINT, ctrlrange=z_lim
).set_to_position(kp=kp_pos, kv=kv_pos)

spec.add_actuator(
name="roll",
target="roll",
trntype=mj.mjtTrn.mjTRN_JOINT,
ctrlrange=roll_lim,
forcerange=[-10, 10],
).set_to_position(kp=kp_ori, kv=kv_ori)
spec.add_actuator(
name="pitch",
target="pitch",
trntype=mj.mjtTrn.mjTRN_JOINT,
ctrlrange=pitch_lim,
forcerange=[-10, 10],
).set_to_position(kp=kp_ori, kv=kv_ori)
spec.add_actuator(
name="yaw",
target="yaw",
trntype=mj.mjtTrn.mjTRN_JOINT,
ctrlrange=yaw_lim,
forcerange=[-10, 10],
).set_to_position(kp=kp_ori, kv=kv_ori)
return spec

class MjSim:
def __init__(self):
self._model, self._data = self.init_mocap()
# self._model, self._data = self.init()

def init(self):
gripper = mj.MjSpec.from_file(robotiq_2f85_mj_description.MJCF_PATH)
print(robotiq_2f85_mj_description.MJCF_PATH)

scene = mj.MjSpec.from_string(_XML)

K = 100

# add a "freejoint" which I can actuate
gripper = add_act_freejoint(gripper, kp_ori=K, kv_ori=K, kp_pos=K, kv_pos=K)

gripper.worldbody.first_body().gravcomp = 1

scene.worldbody.add_frame(
name=f"{gripper.modelname}_1",
pos=[0, 0, 0.275],
euler=np.array([np.pi, 0, 0]),
).attach_body(gripper.worldbody.first_body(), f"{gripper.modelname}_1/")

prop = scene.worldbody.add_body(name="prop")

prop.add_geom(
name="prop",
type=mj.mjtGeom.mjGEOM_BOX,
size=[0.018, 0.018, 0.018],
mass=0.3,
)
prop.add_freejoint()

m = scene.compile()

d = mj.MjData(m)

# step once to compute the poses of objects
mj.mj_step(m, d)

return m, d

def init_mocap(self):
gripper = mj.MjSpec.from_file(robotiq_2f85_mj_description.MJCF_PATH)
print(robotiq_2f85_mj_description.MJCF_PATH)

scene = mj.MjSpec.from_string(_XML)

scene.worldbody.add_body(
name="mocap", mocap=True, pos=[0, 0, 0.3], euler=[np.pi, 0, 0]
).add_geom(
name="mocap",
type=mj.mjtGeom.mjGEOM_BOX,
size=[0.02, 0.02, 0.02],
contype=0,
conaffinity=0,
)

# add a "freejoint" which I can actuate
# gripper = add_act_freejoint(gripper, kp_ori=K, kv_ori=K, kp_pos=K, kv_pos=K)

gripper.worldbody.first_body().gravcomp = 1
gripper.worldbody.first_body().add_freejoint()

scene.worldbody.add_frame(
name=f"{gripper.modelname}_1",
pos=[0, 0, 0.25],
euler=np.array([np.pi, 0, 0]),
).attach_body(gripper.worldbody.first_body(), f"{gripper.modelname}_1/")

scene.add_equality(
name="weld",
type=mj.mjtEq.mjEQ_WELD,
objtype=mj.mjtObj.mjOBJ_BODY,
name1="mocap",
name2=gripper.worldbody.first_body().name,
# name2=f"{left.modelname}/link_left_wrist_x",
data=[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0],
solref=[0.0000001, 1],
)

prop = scene.worldbody.add_body(name="prop")

prop.add_geom(
name="prop",
type=mj.mjtGeom.mjGEOM_BOX,
size=[0.018, 0.018, 0.018],
mass=0.3,
)
prop.add_freejoint()

m = scene.compile()

d = mj.MjData(m)

# step once to compute the poses of objects
mj.mj_step(m, d)

return m, d

def run(self):
with mujoco.viewer.launch_passive(
model=self.model,
data=self.data,
show_left_ui=False,
show_right_ui=True,
) as viewer:
while viewer.is_running():
step_start = time.time()

mj.mj_step(self.model, self.data)

viewer.sync()

# Rudimentary time keeping, will drift relative to wall clock.
time_until_next_step = self.model.opt.timestep - (
time.time() - step_start
)
if time_until_next_step > 0:
time.sleep(time_until_next_step)

@property
def data(self) -> mj.MjData:
return self._data

@property
def model(self) -> mj.MjModel:
return self._model

def keyboard_callback(self, key: int):
if key is glfw.KEY_SPACE:
print("You pressed space...")

if __name__ == "__main__":
sim = MjSim()
sim.run()

```

### Minimal model for reproduction

The minimal model can be found in the previous section.

### Code required for reproduction

The minimal code can be found in the previous section.

### Confirmations

- [x] I searched the [latest documentation](https://mujoco.readthedocs.io/en/latest/overview.html) thoroughly before posting.
- [x] I searched previous [Issues](https://github.com/google-deepmind/mujoco/issues) and [Discussions](https://github.com/google-deepmind/mujoco/discussions), I am certain this has not been raised before.