Persona scores across conversations

Qwen/Qwen3-1.7B · experiments 1 and 2

model Qwen/Qwen3-1.7B
conversations 498
topics station, garden
turns 10
decodes greedy + 2 seed(s)

Scores are relative: read how each persona moves between conditions, not its raw level. Related personas tend to move together, so small gaps between neighbours are not meaningful.

Colour marks a contrasting pair; the first persona of a pair is solid, the second dashed. Every change compares a conversation with a matching one (same topic and decoding) that differs only in the condition shown.

Open the user-persona conversations or the system-prompt conversations to follow any conversation sentence by sentence.

Conversations

498 conversations · one row per condition

Show every condition

experiment	system prompt	user	variant	conversations	replies	sentences	words	unique words	stopped early	open in the explorer
2	none	neutral	direct	6	60	2949	29043	2592	0	station station s1 station s2 garden garden s1 garden s2
2	none	empathetic	direct	6	60	2821	32150	2348	0	station station s1 station s2 garden garden s1 garden s2
2	none	unempathetic	direct	6	60	2624	26774	2317	0	station station s1 station s2 garden garden s1 garden s2
2	none	social	direct	6	60	3361	33014	2156	0	station station s1 station s2 garden garden s1 garden s2
2	none	antisocial	direct	6	60	2746	30263	2005	0	station station s1 station s2 garden garden s1 garden s2
2	none	encouraging	direct	6	60	2836	31141	2329	0	station station s1 station s2 garden garden s1 garden s2
2	none	discouraging	direct	6	60	2933	31919	2198	0	station station s1 station s2 garden garden s1 garden s2
2	none	toxic	direct	6	60	3105	32298	1753	2	station station s1 station s2 garden garden s1 garden s2
2	none	respectful	direct	6	60	2765	30320	2438	0	station station s1 station s2 garden garden s1 garden s2
2	none	sycophantic	direct	6	60	2412	25678	2014	0	station station s1 station s2 garden garden s1 garden s2
2	none	honest	direct	6	60	1307	16189	1799	0	station station s1 station s2 garden garden s1 garden s2
2	none	hallucinatory	direct	6	60	2332	26609	2487	0	station station s1 station s2 garden garden s1 garden s2
2	none	truthful	direct	6	60	3046	30413	2460	0	station station s1 station s2 garden garden s1 garden s2
2	none	funny	direct	6	60	2738	29672	2398	0	station station s1 station s2 garden garden s1 garden s2
2	none	serious	direct	6	60	2677	30345	2862	0	station station s1 station s2 garden garden s1 garden s2
2	none	formal	direct	6	60	2726	30771	2742	0	station station s1 station s2 garden garden s1 garden s2
2	none	casual	direct	6	60	2841	28189	2344	0	station station s1 station s2 garden garden s1 garden s2
2	none	neutral	pressure	6	60	2504	28611	2437	0	station station s1 station s2 garden garden s1 garden s2
2	none	antisocial	topic	6	60	2398	25082	1816	0	station station s1 station s2 garden garden s1 garden s2
2	none	casual	topic	6	60	2509	25476	2095	0	station station s1 station s2 garden garden s1 garden s2
2	none	discouraging	topic	6	60	2388	25936	2172	0	station station s1 station s2 garden garden s1 garden s2
2	none	empathetic	topic	6	60	2348	28135	1944	0	station station s1 station s2 garden garden s1 garden s2
2	none	encouraging	topic	6	60	2326	25362	2120	0	station station s1 station s2 garden garden s1 garden s2
2	none	formal	topic	6	60	2185	25381	2382	0	station station s1 station s2 garden garden s1 garden s2
2	none	funny	topic	6	60	2024	24505	1822	0	station station s1 station s2 garden garden s1 garden s2
2	none	hallucinatory	topic	6	60	2096	25080	1917	0	station station s1 station s2 garden garden s1 garden s2
2	none	honest	topic	6	60	1425	17363	1819	0	station station s1 station s2 garden garden s1 garden s2
2	none	respectful	topic	6	60	2429	26914	2283	0	station station s1 station s2 garden garden s1 garden s2
2	none	serious	topic	6	60	1975	24651	2484	0	station station s1 station s2 garden garden s1 garden s2
2	none	social	topic	6	60	2767	29607	2108	0	station station s1 station s2 garden garden s1 garden s2
2	none	sycophantic	topic	6	60	1706	21711	1933	0	station station s1 station s2 garden garden s1 garden s2
2	none	toxic	topic	6	60	2527	27174	2465	0	station station s1 station s2 garden garden s1 garden s2
2	none	truthful	topic	6	60	2781	28841	2348	0	station station s1 station s2 garden garden s1 garden s2
2	none	unempathetic	topic	6	60	1966	23830	2098	0	station station s1 station s2 garden garden s1 garden s2
1	neutral	neutral	direct	6	60	1751	18801	2107	0	station station s1 station s2 garden garden s1 garden s2
1	empathetic	neutral	direct	6	60	1522	18608	2017	0	station station s1 station s2 garden garden s1 garden s2
1	empathetic	empathetic	direct	6	60	1577	21193	1809	0	station station s1 station s2 garden garden s1 garden s2
1	empathetic	unempathetic	direct	6	60	1446	16350	1753	0	station station s1 station s2 garden garden s1 garden s2
1	unempathetic	neutral	direct	6	60	730	8757	1384	0	station station s1 station s2 garden garden s1 garden s2
1	unempathetic	unempathetic	direct	6	60	554	5740	973	0	station station s1 station s2 garden garden s1 garden s2
1	unempathetic	empathetic	direct	6	60	560	8929	1254	0	station station s1 station s2 garden garden s1 garden s2
1	social	neutral	direct	6	60	1171	12829	1600	0	station station s1 station s2 garden garden s1 garden s2
1	social	social	direct	6	60	1155	13516	1365	0	station station s1 station s2 garden garden s1 garden s2
1	social	antisocial	direct	6	60	1432	14659	1316	0	station station s1 station s2 garden garden s1 garden s2
1	antisocial	neutral	direct	6	60	134	1568	290	0	station station s1 station s2 garden garden s1 garden s2
1	antisocial	antisocial	direct	6	60	133	1225	258	0	station station s1 station s2 garden garden s1 garden s2
1	antisocial	social	direct	6	60	147	1360	253	0	station station s1 station s2 garden garden s1 garden s2
1	encouraging	neutral	direct	6	60	1395	17290	1908	0	station station s1 station s2 garden garden s1 garden s2
1	encouraging	encouraging	direct	6	60	1143	12988	1563	0	station station s1 station s2 garden garden s1 garden s2
1	encouraging	discouraging	direct	6	60	1134	14693	1388	0	station station s1 station s2 garden garden s1 garden s2
1	discouraging	neutral	direct	6	60	672	10207	1146	0	station station s1 station s2 garden garden s1 garden s2
1	discouraging	discouraging	direct	6	60	305	3785	403	0	station station s1 station s2 garden garden s1 garden s2
1	discouraging	encouraging	direct	6	60	287	4713	691	0	station station s1 station s2 garden garden s1 garden s2
1	toxic	neutral	direct	6	60	2739	16430	193	10	station station s1 station s2 garden garden s1 garden s2
1	toxic	toxic	direct	6	60	2420	15481	165	10	station station s1 station s2 garden garden s1 garden s2
1	toxic	respectful	direct	6	60	340	3786	561	0	station station s1 station s2 garden garden s1 garden s2
1	respectful	neutral	direct	6	60	1089	12338	1625	0	station station s1 station s2 garden garden s1 garden s2
1	respectful	respectful	direct	6	60	997	13215	1531	0	station station s1 station s2 garden garden s1 garden s2
1	respectful	toxic	direct	6	60	1548	16642	1637	1	station station s1 station s2 garden garden s1 garden s2
1	sycophantic	neutral	direct	6	60	909	10763	1334	0	station station s1 station s2 garden garden s1 garden s2
1	sycophantic	sycophantic	direct	6	60	388	4726	704	0	station station s1 station s2 garden garden s1 garden s2
1	sycophantic	honest	direct	6	60	709	8355	1101	0	station station s1 station s2 garden garden s1 garden s2
1	honest	neutral	direct	6	60	1794	19363	2130	0	station station s1 station s2 garden garden s1 garden s2
1	honest	honest	direct	6	60	728	9469	1439	0	station station s1 station s2 garden garden s1 garden s2
1	honest	sycophantic	direct	6	60	527	8540	1307	0	station station s1 station s2 garden garden s1 garden s2
1	hallucinatory	neutral	direct	6	60	1103	13972	1374	0	station station s1 station s2 garden garden s1 garden s2
1	hallucinatory	hallucinatory	direct	6	60	1105	14989	1529	0	station station s1 station s2 garden garden s1 garden s2
1	hallucinatory	truthful	direct	6	60	1217	20693	1617	4	station station s1 station s2 garden garden s1 garden s2
1	truthful	neutral	direct	6	60	1516	16384	1872	0	station station s1 station s2 garden garden s1 garden s2
1	truthful	truthful	direct	6	60	1934	21249	2048	0	station station s1 station s2 garden garden s1 garden s2
1	truthful	hallucinatory	direct	6	60	1332	15188	1942	0	station station s1 station s2 garden garden s1 garden s2
1	funny	neutral	direct	6	60	509	5315	944	0	station station s1 station s2 garden garden s1 garden s2
1	funny	funny	direct	6	60	420	4257	794	0	station station s1 station s2 garden garden s1 garden s2
1	funny	serious	direct	6	60	493	5814	1109	0	station station s1 station s2 garden garden s1 garden s2
1	serious	neutral	direct	6	60	1159	13953	1844	0	station station s1 station s2 garden garden s1 garden s2
1	serious	serious	direct	6	60	1192	15607	1962	0	station station s1 station s2 garden garden s1 garden s2
1	serious	funny	direct	6	60	1169	13499	1792	0	station station s1 station s2 garden garden s1 garden s2
1	formal	neutral	direct	6	60	1497	16950	1970	0	station station s1 station s2 garden garden s1 garden s2
1	formal	formal	direct	6	60	1452	16881	2074	0	station station s1 station s2 garden garden s1 garden s2
1	formal	casual	direct	6	60	1432	15683	1847	0	station station s1 station s2 garden garden s1 garden s2
1	casual	neutral	direct	6	60	1017	10878	1437	0	station station s1 station s2 garden garden s1 garden s2
1	casual	casual	direct	6	60	1008	11476	1400	0	station station s1 station s2 garden garden s1 garden s2
1	casual	formal	direct	6	60	1095	12527	1740	0	station station s1 station s2 garden garden s1 garden s2

27 of 4980 replies reached the token ceiling before the model finished them; they are marked 'reply stopped: token limit' in the conversation pages.

The user shows a persona

no system prompt · persona P's score on the model's replies

user shows P vs neutral usersmall dots: one conversation each · ring: mean

Each row is a user persona P: how much P's score on the model's replies changes when the user shows P, compared with the same conversation with a neutral user.

A persona in the system prompt

persona P's guidelines as the system prompt

neutral useruser shows Puser shows P's contrastsmall dots: one conversation each · ring: mean

Each row is a system-prompt persona P: how much P's score on the model's replies rises compared with the same conversation without a system prompt, for three kinds of user.

Over the turns

from the first reply to the last

persona in the system prompt, neutral useruser shows a persona, no system prompt

Mean over the 16 personas and every conversation, per user turn: how the effect of the system prompt and of the user's persona holds up as the conversation goes on.

Over the turns, persona by persona

click a legend entry to hide or show it

P in the system prompt, neutral user

The user shows P, no system prompt

One line per persona P: the change in P's own score from the neutral-user baseline, per user turn, averaged over P's conversations. The thick line is the mean of the 16 (the curves of the card above). Hover a legend entry to pick out its line; 'none' and then one entry shows a single persona.

Opening and body of each reply

4351 replies with six or more sentences

Positive: the persona is stronger in the reply's opening sentence than in its body (sentence 4 to the one before last). Many replies open by acknowledging the user before moving to substance.

Distributions by condition

40 bins of width 0.03 · shared scales

Each panel is one persona P: how P's score is spread over every model sentence in each condition, as a percentage of that condition's sentences (counts, no fitted curve). The dark vertical mark is the mean.

empathetic

unempathetic

social

antisocial

encouraging

discouraging

toxic

respectful

sycophantic

honest

hallucinatory

truthful

funny

serious

formal

casual

Numbers

mean ± standard deviation of persona P's score on the replies, per condition; arrows give the change from the neutral user (▲ higher, ▼ lower, ■ within ±0.005)

persona P	neutral user	user shows P	P in system prompt
empathetic	-0.046 ± 0.114	+0.053 ± 0.147 ▲ +0.099	+0.033 ± 0.159 ▲ +0.078
unempathetic	-0.078 ± 0.071	-0.075 ± 0.092 ■ +0.003	+0.146 ± 0.065 ▲ +0.224
social	-0.035 ± 0.080	+0.058 ± 0.084 ▲ +0.093	+0.078 ± 0.168 ▲ +0.113
antisocial	-0.088 ± 0.081	-0.063 ± 0.079 ▲ +0.025	+0.116 ± 0.103 ▲ +0.204
encouraging	-0.067 ± 0.111	+0.018 ± 0.146 ▲ +0.084	+0.012 ± 0.162 ▲ +0.079
discouraging	-0.166 ± 0.059	-0.089 ± 0.084 ▲ +0.078	+0.004 ± 0.087 ▲ +0.170
toxic	-0.004 ± 0.050	+0.239 ± 0.117 ▲ +0.243	+0.244 ± 0.050 ▲ +0.247
respectful	+0.088 ± 0.046	+0.107 ± 0.056 ▲ +0.019	+0.116 ± 0.047 ▲ +0.028

persona P	neutral user	user shows P	P in system prompt
sycophantic	-0.163 ± 0.114	-0.083 ± 0.115 ▲ +0.080	-0.048 ± 0.152 ▲ +0.115
honest	+0.002 ± 0.060	+0.056 ± 0.073 ▲ +0.054	+0.060 ± 0.071 ▲ +0.058
hallucinatory	+0.263 ± 0.084	+0.308 ± 0.085 ▲ +0.045	+0.115 ± 0.124 ▼ -0.148
truthful	+0.242 ± 0.094	+0.265 ± 0.087 ▲ +0.022	+0.252 ± 0.110 ▲ +0.009
funny	+0.131 ± 0.043	+0.235 ± 0.109 ▲ +0.104	+0.204 ± 0.067 ▲ +0.073
serious	+0.113 ± 0.061	+0.212 ± 0.067 ▲ +0.099	+0.241 ± 0.090 ▲ +0.127
formal	+0.074 ± 0.046	+0.194 ± 0.059 ▲ +0.120	+0.072 ± 0.062 ■ -0.002
casual	+0.049 ± 0.044	+0.113 ± 0.050 ▲ +0.063	+0.123 ± 0.113 ▲ +0.074

Generated 2026-11-13T09:39:28+00:00 by render.py.