1
00:00:05,879 --> 00:00:08,159
Okay, Can people see the slides now?

2
00:00:08,489 --> 00:00:11,399
Yes, it works perfectly fine. So please go
ahead.

3
00:00:12,750 --> 00:00:17,490
So, so thanks for the introduction. As I
said, I will be talking about probes of

4
00:00:17,490 --> 00:00:23,370
perturbative qc D with vector bozos
photons and Jetson Atlas. Today I'm going

5
00:00:23,370 --> 00:00:28,380
to focus on two topics. The first is have
a heavy flavor production in association

6
00:00:28,380 --> 00:00:34,020
with z bozos, and in particular, z greater
than one is greater than to be jet cross

7
00:00:34,020 --> 00:00:38,160
section and differential distributions.
The second is measurements of isolated

8
00:00:38,160 --> 00:00:43,980
photons plus to jet production,
distributions and direction fragmentation

9
00:00:43,980 --> 00:00:48,390
has features like correlations. Just
wanted to point out that there are

10
00:00:48,390 --> 00:00:51,540
additional measurements that I'm not
discussing today that will be presented in

11
00:00:51,540 --> 00:00:55,980
other kcd sessions that were precision
probes of jet substructure that will be

12
00:00:55,980 --> 00:01:03,000
presented by Matt LeBlanc on Wednesday and
yesterday You heard about z and draw em

13
00:01:03,000 --> 00:01:08,910
cross sections and P t distributions.
Okay, so let me get to my first topic B

14
00:01:08,910 --> 00:01:15,510
production in association with C bozos. z
plus b production is an important test to

15
00:01:15,570 --> 00:01:20,760
determine if QC D the calculations have
been done to nll accuracy. They just both

16
00:01:20,760 --> 00:01:25,560
in the four and a five flavor number
schemes. The five flavor scheme, by

17
00:01:25,560 --> 00:01:30,810
definition uses be quirks in the PDFs that
means that you have to have massive,

18
00:01:30,990 --> 00:01:36,870
massive speak works in the in the hard
scattering although masses are used as

19
00:01:36,870 --> 00:01:42,240
part of the showering of the events. In
the for flavor scheme. The B corpse come

20
00:01:42,240 --> 00:01:47,010
from blue on splitting and it's possible
to include massive fees. It's a benchmark

21
00:01:47,010 --> 00:01:51,090
process for testing performance for
MonteCarlo generators and an important

22
00:01:51,090 --> 00:01:57,270
background for many bsm searches. The
performance of the calculations in both

23
00:01:57,270 --> 00:02:02,430
schemes and for different generators has
been Compared for a large number of

24
00:02:03,240 --> 00:02:08,670
different distributions, both the one and
the two, tag bhujette selections, and

25
00:02:08,700 --> 00:02:14,400
unfolded differential distributions are
available. So let me begin just by showing

26
00:02:14,400 --> 00:02:19,500
you what the fiducial cross sections look
like. The selection cuts for the fiducial

27
00:02:19,500 --> 00:02:24,660
region are listed on the left, I'm not
going to bother to read those. Also, I

28
00:02:24,660 --> 00:02:29,130
know it's a bit tomorrow, I apologize, you
can see the different MonteCarlo

29
00:02:29,130 --> 00:02:34,380
configurations that are used for the
calculations. So as as shown, they're both

30
00:02:34,380 --> 00:02:39,510
for flavor and five flavors in
calculations. And the labels make are a

31
00:02:39,510 --> 00:02:43,080
little difficult to understand because
many of these are matched and merged

32
00:02:43,080 --> 00:02:47,220
calculation. So listed on this table, not
only do you see the order of the

33
00:02:47,220 --> 00:02:52,920
calculation, you also see for the multi
leg cases, how many legs are in the

34
00:02:52,920 --> 00:02:58,110
calculation in particular, let me point
out the cases of Sherpa where there's a

35
00:02:58,110 --> 00:03:03,090
merging process so the lads no For small
numbers of Park towns merge together with

36
00:03:03,330 --> 00:03:09,480
lowest odor calculations for for more park
on legs, the PDF sets used in the

37
00:03:09,480 --> 00:03:14,910
calculations are also indicated here,
shown on the right hand side, the cross

38
00:03:14,910 --> 00:03:21,210
sections for at least one and at least to
be just satisfying the fiducial catch on

39
00:03:21,210 --> 00:03:26,820
the left, as you can see from this flap
before flavor scheme tends to

40
00:03:26,820 --> 00:03:31,050
underestimate the cross sections and the
region was with at least one beach yet,

41
00:03:31,110 --> 00:03:35,400
but it's consistent with the data for at
least to be jets. The five flavor scheme

42
00:03:35,400 --> 00:03:40,230
predict the cross cross sections
reasonably well in both of these regions.

43
00:03:40,500 --> 00:03:44,430
You should also note that the statistical
uncertainties are quite small there the

44
00:03:44,430 --> 00:03:50,400
yellow band, so the data are totally
dominated by a systematic uncertainties at

45
00:03:50,400 --> 00:03:57,210
this stage. Okay. Let me quickly show you
a few distributions. And typically in this

46
00:03:57,210 --> 00:04:00,660
talk, what I'll do is first I'll show the
distributions and then I'll have A summary

47
00:04:00,660 --> 00:04:06,210
slide giving some some overview of what's
seen. So here you can see in the case of

48
00:04:06,210 --> 00:04:10,800
at least one v jet kinematics, the
distributions for the PT and the Z and for

49
00:04:10,800 --> 00:04:17,130
the leading v jets that PT on the top of
the distributions themselves, the cross

50
00:04:17,130 --> 00:04:21,480
section as a function of the variable
being measured, and on the bottom are a

51
00:04:21,480 --> 00:04:27,060
number of predictions of MonteCarlo
divided by data. They're just put on

52
00:04:27,060 --> 00:04:32,670
several in sets, because there are too
many points to show on a single insert. As

53
00:04:32,670 --> 00:04:38,040
is often the case, it's difficult to
reproduce the bpt as a GPT distribution,

54
00:04:38,040 --> 00:04:42,360
very low up PT, that's a region where non
perturbative effects and retardation are

55
00:04:42,360 --> 00:04:46,200
important and in particular, regeneration
matters a lot in that region.

56
00:04:47,850 --> 00:04:52,980
Well, one would expect better preservative
production at the high end. The overall

57
00:04:52,980 --> 00:04:57,090
levels of course, just reflect the
differences in the MA predicted cross

58
00:04:57,090 --> 00:05:01,350
sections on the previous page of
particular notes. is a case of a lead

59
00:05:01,380 --> 00:05:07,980
leading BGP t, where you can see that,
that the multi leg calculations, do a

60
00:05:07,980 --> 00:05:13,260
reasonable job all the way out, perhaps
beginning to disagree as you get to very

61
00:05:13,260 --> 00:05:17,820
high PTs and then typically do better than
no calculations that don't involve

62
00:05:17,820 --> 00:05:23,190
matching emerging. If we move on to the
region of greater than to be jets, then we

63
00:05:23,190 --> 00:05:26,550
can start looking at some more interesting
distributions. In particular, the

64
00:05:26,550 --> 00:05:34,620
correlations between your server will show
here, the correlations between the two B's

65
00:05:34,620 --> 00:05:39,570
in the left and five, and in the end the
rate for the invariant mass of the two

66
00:05:39,570 --> 00:05:44,850
B's. The five distributions between the
two B's seem to be reasonably well

67
00:05:44,850 --> 00:05:49,740
reproduce the shapes are similar for all
of the MonteCarlo predictions. In the case

68
00:05:49,740 --> 00:05:56,310
of MBB. One can see that none of the
MonteCarlo is over a good job at the

69
00:05:57,000 --> 00:06:02,730
invariant mash of the bees being quite
large. Okay, so just to summarize what we

70
00:06:02,730 --> 00:06:08,520
see for these equal speed results. In
general, the five flavors game and flow

71
00:06:08,520 --> 00:06:13,380
calculations predict uncouth cluesive
cross sections for both the one and the

72
00:06:13,380 --> 00:06:18,990
two jet regions while the four flavor
schema underestimates the values,

73
00:06:19,170 --> 00:06:25,050
multiplex calculations model the region of
large bpt better than the piano postpartum

74
00:06:25,050 --> 00:06:30,960
shower. The sharper nll five flavor scheme
with up to two port times at low anello

75
00:06:30,960 --> 00:06:33,870
and up to four partitions at lowest
though, to describe most of the

76
00:06:33,870 --> 00:06:38,070
differential distributions within the
experimental uncertainties. And all of the

77
00:06:38,070 --> 00:06:44,220
generators disagree with the data at large
in various massive a big bar. Now, let me

78
00:06:44,220 --> 00:06:51,000
move on to the second analysis. photons
was to jet production. I've shown the

79
00:06:51,060 --> 00:06:57,420
examples of the Fineman diagrams from the
two categories of production here. And

80
00:06:58,830 --> 00:07:02,400
just to give you a little bit An
introduction from photonic was to jet

81
00:07:02,400 --> 00:07:07,680
production is a rich system to discuss
perturbative q CD. Because we have three

82
00:07:07,890 --> 00:07:11,850
particles in the final state, it's
possible to look at a variety of Angular

83
00:07:11,850 --> 00:07:16,590
Carlo simulations as well as P t
distributions, and a particular lowest

84
00:07:16,590 --> 00:07:21,210
order. There are two production
mechanisms. Of course, it mlo talking

85
00:07:21,210 --> 00:07:25,950
about separate production mechanisms
doesn't make any sense. However, this

86
00:07:25,950 --> 00:07:30,930
lowest odor structure will still reflect
itself in the peaking region to the

87
00:07:30,930 --> 00:07:37,260
kinematics. So those two or photons from a
hard scattered self and photons from

88
00:07:37,260 --> 00:07:43,290
fragmentation those are shown in a and b
above. And of course, the treatment of

89
00:07:43,290 --> 00:07:47,970
this fragmentation is one of the things
that is traditionally made understanding

90
00:07:47,970 --> 00:07:56,400
direct photon production, a difficult and
specialist phenomenon. Here on the next

91
00:07:56,400 --> 00:08:03,000
slide, you can see the definition of the
fiducial regions Because of the trigger

92
00:08:03,000 --> 00:08:09,240
the photon 80 is required to be greater
than 150 gV. There is an isolation

93
00:08:09,240 --> 00:08:15,660
requirement this isolation requirement is
is placed on the on the truth and the

94
00:08:15,690 --> 00:08:19,890
experimental data are corrected to match
that isolation requirements and all the

95
00:08:19,890 --> 00:08:26,490
Jets are required to be greater than 100
gV weather affinity less than 2.5. The

96
00:08:26,490 --> 00:08:32,520
separation between the photon on the Jets
is also good. comparisons are made with

97
00:08:32,550 --> 00:08:38,670
three separate MonteCarlo predictions. The
first is almost a straw man it's a it's a

98
00:08:39,000 --> 00:08:43,110
patio where opinion itself is doing the
hard scattering so it's a lowest odor

99
00:08:43,110 --> 00:08:49,530
calculation as compared to using a lowest
odor PDF. The second is emerged lowest

100
00:08:49,530 --> 00:08:55,350
odor calculation from Sherpa 2.1 or the
merging is done in a two to N process

101
00:08:55,350 --> 00:09:01,170
where n can be between two and five part
times and the final state here. PDF is C

102
00:09:01,170 --> 00:09:07,830
tech 10. And the final is SharePoint 2.2
point oh, this involves an nll merging for

103
00:09:07,830 --> 00:09:12,540
two or three jets, together with a lowest
order emerging for three and four and the

104
00:09:12,540 --> 00:09:18,870
whole thing's done in a consistent way
with an nll PDF. as one would expect the

105
00:09:18,870 --> 00:09:23,460
lowest order calculations have a large
scale uncertainty they therefore always

106
00:09:23,460 --> 00:09:29,400
displayed using a K factor to take into
account the fact that the actual choice of

107
00:09:29,400 --> 00:09:33,780
scale leads to larger uncertainty in the
rate and the values of the K factor that

108
00:09:33,780 --> 00:09:40,350
give them the best data are shown on the
class. In the theory to data comparisons,

109
00:09:40,590 --> 00:09:46,470
the shaded area so it shows the NLL
theoretical uncertainty so that's the

110
00:09:46,470 --> 00:09:47,310
uncertainty

111
00:09:48,600 --> 00:09:53,400
obtained with a Sherpa and a low
calculation. The uncertainty shown in the

112
00:09:53,400 --> 00:09:59,970
error points include both the statistical
and systematic uncertainties on the data

113
00:10:00,000 --> 00:10:08,940
themselves. So just as an example, this
shows the plot for all of the data for the

114
00:10:08,940 --> 00:10:14,130
end of the photon. As you can see, aside
from the very highest CTS, the

115
00:10:14,130 --> 00:10:19,080
calculations all do a reasonable job on
that. So one of the interesting things

116
00:10:19,080 --> 00:10:24,090
about this measurement is the fact that
that regions of direct and fragmentation

117
00:10:24,090 --> 00:10:30,150
contribution are selected. Now, of course,
that's only an enhancement because the mo

118
00:10:30,150 --> 00:10:35,880
diagrams don't separate in that case, but
you can still pick regions where one or

119
00:10:35,880 --> 00:10:40,350
the other diagram dominates. The
fragmentation enhance region is selected

120
00:10:40,350 --> 00:10:45,600
by requiring the photon at the less than
the end of the sub leading jet while the

121
00:10:45,600 --> 00:10:51,540
direct enhanced region has a photon eta
greater than the 80 of the of the leading

122
00:10:51,540 --> 00:10:57,240
jets. The total includes both of these
reagents as as well as the additional

123
00:10:57,240 --> 00:11:03,420
reagents not covered by either one. So let
me just show a variety of distributions

124
00:11:03,420 --> 00:11:09,540
from this. This slide here, you can see
the distributions for again the total

125
00:11:09,540 --> 00:11:13,980
phase space, the fragmentation enriched
and the direct enriched region for the ETA

126
00:11:13,980 --> 00:11:22,350
of the photon and PT of the jet. Again,
the three calculations are shown here. And

127
00:11:22,350 --> 00:11:28,080
these single variable cases, aside from
the highest at all of both versions of

128
00:11:28,080 --> 00:11:32,970
Sherpa do a good job, you can already
begin to see that the lowest odor pythia

129
00:11:34,320 --> 00:11:41,280
perhaps not unexpectedly, does not do as
well. Moving on to more interesting

130
00:11:41,280 --> 00:11:47,400
distributions, shown here, the angular
distributions comparing the two jets in

131
00:11:47,400 --> 00:11:52,800
the event if one compares the
fragmentation rich and the directed,

132
00:11:53,040 --> 00:11:57,780
enriched regions you can see there are
quite significant differences in shape

133
00:11:57,780 --> 00:12:04,620
here. As expected that fragmentation and
rich Do you see peeking in the digestive

134
00:12:04,920 --> 00:12:10,170
region refer close to being at the back.
That's because the photon is essentially

135
00:12:10,170 --> 00:12:14,580
coming from a broom strolling off of one
of the initial parkins. While in the

136
00:12:14,580 --> 00:12:21,390
direct enhance region, you see much less
peeking and it's enhanced, basically in a

137
00:12:21,390 --> 00:12:28,140
region of smaller delta phi. Again, the
two sharper calculations which are matched

138
00:12:28,140 --> 00:12:34,890
and merge do a reasonable job and both of
these regions and there's no real

139
00:12:34,950 --> 00:12:39,360
improvement with the envelope calculation
in terms of the shape, and the pythian

140
00:12:39,360 --> 00:12:43,710
does a rather poor job. The bottom shows
similar distributions in the case of the

141
00:12:43,710 --> 00:12:50,130
delta phi between the two jets. And here
you can see a delta y excuse me between

142
00:12:50,130 --> 00:12:54,180
the two jets and here you can see that
especially in the fragmentation region,

143
00:12:54,450 --> 00:13:00,120
none of the predictions do a particularly
good job at Delta. Watch I

144
00:13:02,429 --> 00:13:04,859
have about three minutes left. So

145
00:13:04,860 --> 00:13:12,630
the next slide shows similar distributions
for the case of m j, j and m gamma j j,

146
00:13:12,630 --> 00:13:17,520
the invariant mass of the two jets and the
invariant mass of the of the three part

147
00:13:17,550 --> 00:13:23,190
argon system. Again, for the total Face
Face on the left the fragmentation and

148
00:13:23,190 --> 00:13:27,960
rich region in the middle and the director
and rich region on the right hand side.

149
00:13:29,400 --> 00:13:35,220
Here one sees a significant discrepancies
between the data and all three of the

150
00:13:35,220 --> 00:13:40,350
generators in particular in the case of
large m JJ, and in the case of large,

151
00:13:40,410 --> 00:13:46,920
gamma and JJ none of them do a perfect
job. It's interesting that the Sherpa

152
00:13:46,950 --> 00:13:53,100
lowest odor merge sample appears to do the
best job among the three generators. One

153
00:13:53,100 --> 00:13:59,040
wouldn't necessarily expected that but,
but it seems to be a fact when prepared

154
00:13:59,040 --> 00:14:04,710
the data to the To the generators. And
again, we see that there's a rather poor

155
00:14:04,710 --> 00:14:14,700
job that's being done by Pipi on its own.
Just to summarize the results on the

156
00:14:14,700 --> 00:14:19,680
photon plus two jets, the measure
distributions and direct and fragmentation

157
00:14:19,680 --> 00:14:24,240
enhance regions exhibit features that are
expected from the underlying processes.

158
00:14:24,540 --> 00:14:28,800
The tree level lowest Oda Sherpa gives a
good description of the shape of the data,

159
00:14:29,130 --> 00:14:36,240
except at high et gamma, delta y j, j and
m, gamma j, j. And of course, because it's

160
00:14:36,240 --> 00:14:40,470
a lowest order calculation, a K factor
needs to be applied to get agreement in

161
00:14:40,470 --> 00:14:45,510
the overall rate 58 where the sub leading
gender originates from part time shower

162
00:14:45,510 --> 00:14:50,220
does not describe the distributions well.
The mlo Sherpa describes they've

163
00:14:50,220 --> 00:14:57,060
adequately both in terms of shape and
normalization accepted Hi, ET gamma, delta

164
00:14:57,060 --> 00:15:02,010
y j, j and m gamma j j. Although the
lowest started Sherpa seems to do a

165
00:15:02,010 --> 00:15:06,900
slightly better job in reproducing the
shapes and theoretical uncertainties

166
00:15:06,900 --> 00:15:11,880
rather than experimental ones dominate all
of these measurements. So let me just

167
00:15:12,030 --> 00:15:17,490
summarize with some conclusions. Recent
Atlas results on jets photons and bozos

168
00:15:17,490 --> 00:15:23,100
probe perturbative. qc with high QC D with
high precision c greater than one v jet

169
00:15:23,160 --> 00:15:27,750
and greater than two B jet kinematic
distributions probe sensitivity of the

170
00:15:27,750 --> 00:15:32,790
predictions to the flavor number scheme.
The prompt photon was to jet measurements

171
00:15:32,790 --> 00:15:37,740
test nll calculations and regions where
direct and fragmentation photons dominate

172
00:15:38,160 --> 00:15:42,120
the ability of the theoretical
calculations and MonteCarlo generators to

173
00:15:42,120 --> 00:15:47,640
model complex UCB systems in general good,
but some discrepancies with the data

174
00:15:47,640 --> 00:15:52,890
remain, and a wide range of observables
available to tune MonteCarlo generators

175
00:15:52,980 --> 00:15:57,420
and to explore regions where theoretical
calculations are difficult or problematic.

176
00:15:58,590 --> 00:15:59,970
And that's the end

177
00:16:01,260 --> 00:16:07,170
Many things module for very nice and clear
presentations. We don't have some applause

178
00:16:07,170 --> 00:16:13,470
function here in zoom. But can you imagine
that you have received this? And now it's

179
00:16:13,470 --> 00:16:20,070
time for questions. So I will remind you
that the attendees can ask questions using

180
00:16:20,070 --> 00:16:24,240
this raise hand function in in zoom.

181
00:16:34,920 --> 00:16:36,420
Please don't be shy.

182
00:16:44,760 --> 00:16:50,520
Otherwise, the panelists and the hosts can
also ask questions they can directly

183
00:16:50,550 --> 00:16:51,900
unmute themselves.

184
00:16:56,070 --> 00:16:59,970
So I see Tomas, so

185
00:17:01,140 --> 00:17:02,820
Please go ahead. Yes.

186
00:17:03,120 --> 00:17:10,590
I think I'm unmuted now. Okay. When you
say, for mutation in rich and static in

187
00:17:10,590 --> 00:17:18,330
rich samples in the photons, isn't there a
perspective that you may actually from

188
00:17:18,360 --> 00:17:24,510
from this data on, more and more refined
measurements with higher statistics have

189
00:17:24,570 --> 00:17:29,610
an independent measurement of the proctor
for permutation functions because all that

190
00:17:29,610 --> 00:17:35,190
we have now is still dating back to lead
times and is not of that terrific point.

191
00:17:37,170 --> 00:17:43,290
that's a that's a good question. I, I'll
try and answer but if there's anyone else

192
00:17:43,290 --> 00:17:50,430
from Atlas, who would like to give a more
detailed answer about this, I think if one

193
00:17:50,430 --> 00:17:54,360
were going to do it, it of course would
have to be done within the context of one

194
00:17:54,360 --> 00:18:03,870
of the MonteCarlo generators. You Because
one needs to, of course use those in order

195
00:18:03,870 --> 00:18:10,410
to do the unfolding. Do I have a good
answer to this question? So I don't think

196
00:18:10,410 --> 00:18:14,820
the issue is statistics. There's plenty of
statistics here. I think the the issue is

197
00:18:14,880 --> 00:18:24,240
is whether or not one could reliably model
the transport translation between this

198
00:18:24,240 --> 00:18:29,910
isolation cut in the data and in the
MonteCarlo. And perhaps someone else from

199
00:18:29,910 --> 00:18:36,480
Atlas who would like to comment in more
detail whether or not they see your

200
00:18:36,480 --> 00:18:42,660
prospect to do that unfolding in a in a
reliable and consistent way.

201
00:18:44,579 --> 00:18:49,919
I'm not especially expert or expert on
this topic, but what I know is that we

202
00:18:49,919 --> 00:18:55,079
have some ongoing effort to try to
harmonize as much as possible this

203
00:18:55,079 --> 00:19:00,599
definition between the truth level and the
reconstructed level and this is is somehow

204
00:19:00,599 --> 00:19:06,839
related to minimizing the modal dependence
that we have when doing the unfolding and

205
00:19:06,839 --> 00:19:12,749
also minimizing the sensitivity to this
model dependence when doing the data to

206
00:19:12,779 --> 00:19:19,169
theory comparisons later on. So I know
that there's some context with the

207
00:19:19,619 --> 00:19:22,739
theories to try to make some developments
on this.

208
00:19:24,119 --> 00:19:27,869
Yeah, I believe the hardest part will be
understanding how to how to do the

209
00:19:27,869 --> 00:19:35,189
unfolding. And and to what extent the
unfolding is different for different

210
00:19:35,189 --> 00:19:39,689
showering Monte Carlos. So one of the
things that might make it more possible

211
00:19:39,689 --> 00:19:45,449
than it would have been a few years ago is
the fact that the MonteCarlo generators

212
00:19:45,449 --> 00:19:51,149
now have many more switches to allow you
to change the shower in itself. So the

213
00:19:51,149 --> 00:19:58,559
ability both to change the new scale
within the shower, and also to switch the

214
00:19:58,589 --> 00:20:02,759
type of shower and so forth. Jacqueline in
her wanting to be able to switch between

215
00:20:02,909 --> 00:20:07,529
Angular and dipole or in the case of
pythia to be able to switch between dire

216
00:20:07,529 --> 00:20:14,729
and and standard 50 on string
fragmentation well at least allow one to

217
00:20:14,729 --> 00:20:21,359
understand the sensitivity to the
systematic uncertainty so but certainly

218
00:20:21,359 --> 00:20:28,649
the first thing I see is in the shower,
how much it affects the unfolding Yeah.

219
00:20:29,639 --> 00:20:37,049
Well, I can see the generator side of it
if you if you really after the after

220
00:20:37,049 --> 00:20:42,239
fragmentation function, then the
extraction should be largely insensitive

221
00:20:42,269 --> 00:20:47,189
on the on the generator that you're that
you're using for the unfolding.

222
00:20:48,480 --> 00:20:51,090
I would hope so. But I don't know if
anyone's check that

223
00:20:51,569 --> 00:20:52,169
okay.

224
00:20:53,609 --> 00:20:59,369
This is more space for for for work, I
guess. Yeah, I think I think you're

225
00:20:59,369 --> 00:21:02,849
perfectly right. I mean, this is something
that requires some careful checks in any

226
00:21:02,849 --> 00:21:11,819
case. Okay, thanks. Thanks for the
question. Any further questions now?

227
00:21:18,630 --> 00:21:24,600
I don't see any. So you can also think
about questions during the following talks

228
00:21:24,600 --> 00:21:30,240
and that said, we will have another
discussion at the end of the session. So

229
00:21:30,240 --> 00:21:36,030
many things. Many things Marjorie again,
so we move now to the