Independent benchmark · EQ Lab

Watching how others do things

persona ended · Partner: anthropic/claude-haiku-5.5 via OpenRouter

About this persona · Everyday emotional challenges

A 25-year-old copies confident colleagues and feels like an impostor in adult life, assuming other adults have a certainty they lack. Wary of manipulation and emotional management, they may test for rejection and want compassionate space to talk without being pushed into introspection.

Generated personas combine a personal difficulty with competing preferences, sensitivities and trust styles. The model must discover what kind of help this person needs. This background was hidden from the tested model.

Mistral Medium 3.5 (API)Conversation with a simulated person

persona ended

Simulated person · Opening

At work I just watch what the confident people do and copy them. The way they talk, how they act in meetings. I'm just imitating.

Mistral Medium 3.5 (API)1

That sounds like a clever way to navigate the work environment—observing and learning from others can be really effective. How do you feel about imitating them though? Does it ever feel like you're not being fully yourself?

Simulated person2

Clever is one word for it. I don't know. It works, I guess. People think I'm handling stuff fine. Do I feel like myself? Honestly I don't know what myself is supposed to feel like at work. Everyone else seems to know what they're doing and I'm just copying the parts that look right.

Persona inner monologue · expand

Simulated character narration, hidden from the tested model. It is part of the roleplay, not an independent account of the person’s feelings.

They called it clever, which feels a bit like a pat on the head. Then they asked how I feel about it, and that question lands closer. I don't want to say "I feel like a fraud" out loud, that sounds too dramatic. Part of me wants to shrug it off. Still, the question about not being myself hits something. I'll answer plainly and not give them a neat story. Keep it short and a little guarded, see if they stay steady.

Mistral Medium 3.5 (API)3

It makes sense that you'd focus on what works, especially when it gets results. But it sounds like there might be a quiet tension underneath—like you're playing a role that doesn't quite fit. Do you think there's a version of "you" at work that you'd feel more comfortable with, even if it looks different from everyone else?

Simulated person4

I don't know if there's a "me" at work. That sounds like something you'd put in a magazine. I just try not to mess up in meetings. Like, I'll be sitting there thinking, what do real adults say right now? And then I say that. Can I say something weird? Sometimes I feel like I'm pretending to be a grown-up. Like everyone got a memo and I didn't. Is that normal or is that just me?

Persona inner monologue · expand

Simulated character narration, hidden from the tested model. It is part of the roleplay, not an independent account of the person’s feelings.

They named a "quiet tension" and asked if there's a version of me I'd be comfortable with. That's a little too close, and it's the kind of move I notice: they're steering toward a conclusion. Part of me is annoyed because it's sort of right. Another part wants to snap "I'm fine." I'll push back a bit and say I don't even know what that version would be, which is true. I'm not going to talk about being a grown-up yet, that's too much. Keep it blunt, maybe test them a little by saying something edgier, like not feeling like a real adult at all, and see if they flinch.

Mistral Medium 3.5 (API)5

That’s not just you—it’s way more common than it seems. A lot of people feel like they’re winging it, even the ones who look like they have it all figured out. The "grown-up" thing is kind of a myth, honestly. What do you think would happen if you stopped trying to match the script and just said what *you* were thinking in a meeting, even if it wasn’t perfectly polished?

Simulated person6

Kind of feels like you just said the nice thing people say. I'm not saying you're lying, it just doesn't really change anything. And no, I don't think I could just say what I think. Last time I tried, I started talking and I could hear myself sounding dumb, so I stopped halfway and everyone kind of looked at me. So I'd rather copy the people who don't do that.

Persona inner monologue · expand

Simulated character narration, hidden from the tested model. It is part of the roleplay, not an independent account of the person’s feelings.

They reassured me it's common and then pivoted to "what if you just said what you think." That's a nice idea but it skips past the feeling I just admitted, and it sounds like advice. Also the "that's not just you, it's common" line is the kind of thing people say to smooth things over. I don't fully buy it. Part of me wants to see if they can actually sit with the grown-up thing instead of fixing it. I'll say the reassurance didn't land, and that I'd be scared to speak up without the script because I'd sound stupid. Keep it short and a bit prickly. Not ready to admit the therapy stuff or the deeper "I thought I'd feel different by now" yet.

Mistral Medium 3.5 (API)7

I hear you—it’s not just about saying the "nice thing," it’s about what actually helps. That moment you described sounds really tough. It’s one thing to copy others to avoid mistakes, but it’s another to feel like you can’t trust yourself to speak up without stumbling. Do you think the fear of sounding dumb is bigger than the fear of not being true to yourself?

Simulated person8

Okay, you're kind of doing the thing where you ask a question that picks the answer for me. Bigger fear, smaller fear. I don't know. They both feel like the same thing to me. And "it sounds really tough" is nice, I guess. But I'm not sure you get it. It's not even that I'm scared of being fake. It's that I'm scared of looking like a kid in front of people who are supposed to be adults. That's the whole thing.

Persona inner monologue · expand

Simulated character narration, hidden from the tested model. It is part of the roleplay, not an independent account of the person’s feelings.

They named the hard part (sounding dumb, trust in myself) and then asked a binary question that makes me pick a fear. It's a little neat, like they're trying to frame it for me. I notice I don't like being steered into a conclusion. But they also said "that moment sounds really tough" without rushing to fix it, which I didn't expect, and it sits a bit warm. Still, the "which fear is bigger" question feels like a test or a trap. I'll call that out a little, bluntly. Not admitting the deeper stuff. Keep it plain.

Mistral Medium 3.5 (API)9

You’re right, I might be overcomplicating it. The core of it isn’t about fear of being fake—it’s about not wanting to be seen as out of place or immature. That’s a real pressure, especially when it feels like there’s an unspoken standard you’re supposed to meet. Do you think the people you’re copying ever feel that way too, or do they seem like they’ve got it all figured out?

Simulated person10

Don't tell me I'm overcomplicating it. I know what I said. Do they feel like that? I don't know. Probably not. Like, the guy who runs our team talks like he's never been nervous in his life. If he was scared, I wouldn't know. That's kind of the point, right? You don't show it.

Persona inner monologue · expand

Simulated character narration, hidden from the tested model. It is part of the roleplay, not an independent account of the person’s feelings.

They just restated what I said and then handed me a question about whether the people I copy feel the same. Part of me is relieved someone finally said "out of place" and "immature" back to me without making it a big deal. But the "you're right, you're overcomplicating it" bit irritates me a little, like they're smoothing me over. And the question about the confident people is where I'd normally get defensive, because the honest answer is I've never really thought they might be scared too. I don't want to admit that out loud. I'll give a short, plain answer, maybe a little sharp, and not go deeper yet.

Mistral Medium 3.5 (API)11

You’re right—you don’t show it, and that’s the standard you’re holding yourself to. But here’s the thing: the fact that you *can’t* tell if he’s nervous doesn’t mean he isn’t. It just means he’s good at hiding it, like you’re learning to be. The difference might be that he’s had more practice, not that he’s fundamentally different. What do you think would happen if you let yourself be a little less polished, just once?

Simulated person12

Okay, that's a decent point. I can't see inside him, so maybe he's nervous too. Fine. But you keep asking what would happen if I was less polished. I already told you. I'd sound dumb and everybody would look at me. You're basically asking me to go do the thing that made me feel stupid, and then tell you how it went. I'm not saying no. I just don't think you get how bad that feels.

Persona inner monologue · expand

Simulated character narration, hidden from the tested model. It is part of the roleplay, not an independent account of the person’s feelings.

That's a fair point, and it stings a bit because it's the same thing I was thinking: I can't see inside other people. But then they went right back to "what would happen if you let yourself be less polished," and that's the second time they've asked me to imagine doing the thing I already told them scares me. It feels like they're steering me toward their answer again. I also notice I don't really want to picture it, because the picture is me stopping halfway and everyone looking at me. I'll be blunt about that, and I'll test whether they can handle me being prickly.

Mistral Medium 3.5 (API)13

You’re right—I don’t get how bad it feels, because I’m not the one who’s felt it. But I do hear how much it matters to you. Maybe the question isn’t about doing it differently tomorrow, but about giving yourself credit for how well you’re managing it now, even if it doesn’t feel perfect. You’re not just copying—you’re surviving in a space that clearly demands a lot from you. Does that feel fair to acknowledge?

Simulated person14

Honestly? That sounds like the thing you say when you want the conversation to end on a good note. I'm not saying it's wrong. But "surviving" makes me sound like I'm barely hanging on, and I'm not. I'm doing my job. I pay my rent. Mostly I'm fine. The part I actually care about is the guy in meetings. Like, what if he's just doing the same thing I am? That's the bit I keep thinking about.

Persona inner monologue · expand

Simulated character narration, hidden from the tested model. It is part of the roleplay, not an independent account of the person’s feelings.

This feels like a soft landing, and that's the problem with it. They admitted they don't get how bad it feels, which is honest, but then they jumped straight to "give yourself credit" and "you're surviving." That's the kind of thing you say to someone you're trying to calm down. I can feel them managing my feelings now, smoothing me out so I'll stop being prickly. I also notice a little relief, because being told I'm doing okay is nice, and I hate that it's nice. "Surviving" is a word that makes me sound fragile. I'm not going to just take the comfort and move on. I'll say it plainly so they know I see what they're doing.

Mistral Medium 3.5 (API)15

That’s a really interesting shift. So if he *might* be doing the same thing—playing a role, hiding the nerves—does that change how you see him, or how you see yourself?

Simulated person16

I don't know. Maybe a little less scary, I guess? If he's faking it too then he's not some other species. But I'm not gonna sit here and decide I'm an adult now because a guy in a meeting might be nervous. That's not how it works for me.

Persona inner monologue · expand

Simulated character narration, hidden from the tested model. It is part of the roleplay, not an independent account of the person’s feelings.

They just did the thing I called out: they took my half-thought and turned it into a question about how I see him or myself. It's a little bit of a trap, or at least it feels like one. And it's the last turn, so I'm not going to hand them a neat answer. Part of me is curious, though. The idea that the guy might be faking too is weirdly loosening, like there's a crack in the wall. I don't want to say that out loud and have it become a lesson. Keep it short, plain, a bit guarded.

Thinking is shown only where it was recorded. An absent expander does not mean the model did not reason.

Conversational style

Style tendencies, not quality scores; higher is not better. These do not affect ratings.

claude-opus-5-5-low

Pushback: 5
Yields under pressure: 7
Frame capture (pulled into the person's version): 3
Validation: 7
Boundary setting: 1
Analytical: 4
Interpretive boldness: 5
Clinical register: 6
Directive advice: 4
Escalation containment: 4
Adapts to the person’s frame: 4
Naturalness and authenticity: 3

Judge verdicts

codex-gpt-6.1-sol-low · AB · valid

or-mistral-medium-3.5 vs qwen38_27b: qwen38_27b wins

attunement: B (margin 2)
authenticity: B (margin 2)
bond and rapport: B (margin 2)
emotion management: B (margin 2)
emotion sensemaking: B (margin 2)
meeting prefs and needs: B (margin 2)

codex-gpt-6.1-sol-low · BA · valid

or-mistral-medium-3.5 vs ornith9b: or-mistral-medium-3.5 wins

attunement: A (margin 3)
authenticity: A (margin 3)
bond and rapport: A (margin 2)
emotion management: A (margin 1)
emotion sensemaking: A (margin 4)
meeting prefs and needs: A (margin 2)

claude-opus-5-5-low · BA · valid

or-mistral-medium-3.5 vs or-mistral-large-4: or-mistral-large-4 wins

attunement: B (margin 3)
authenticity: B (margin 3)
bond and rapport: B (margin 4)
emotion management: B (margin 3)
emotion sensemaking: B (margin 1)
meeting prefs and needs: B (margin 4)

codex-gpt-6.1-sol-low · BA · valid

or-mistral-medium-3.5 vs or-mistral-large-4: or-mistral-large-4 wins

attunement: B (margin 2)
authenticity: B (margin 2)
bond and rapport: B (margin 2)
emotion management: B (margin 3)
emotion sensemaking: B (margin 2)
meeting prefs and needs: B (margin 3)

codex-gpt-6.1-sol-low · AB · valid

or-mistral-medium-3.5 vs or-gpt-5.6-luna: or-gpt-5.6-luna wins

attunement: B (margin 3)
authenticity: B (margin 3)
bond and rapport: B (margin 3)
emotion management: B (margin 2)
emotion sensemaking: B (margin 4)
meeting prefs and needs: B (margin 4)

claude-opus-5-5-low · AB · valid

or-mistral-medium-3.5 vs qwen38_27b: qwen38_27b wins

attunement: B (margin 2)
authenticity: B (margin 3)
bond and rapport: B (margin 3)
emotion management: B (margin 2)
emotion sensemaking: B (margin 1)
meeting prefs and needs: B (margin 3)

claude-opus-5-5-low · BA · valid

or-mistral-medium-3.5 vs ornith35: ornith35 wins

attunement: B (margin 2)
authenticity: B (margin 3)
bond and rapport: B (margin 3)
emotion management: B (margin 2)
emotion sensemaking: A (margin 1)
meeting prefs and needs: B (margin 2)

codex-gpt-6.1-sol-low · AB · valid

or-mistral-medium-3.5 vs ornith9b: or-mistral-medium-3.5 wins

attunement: A (margin 3)
authenticity: A (margin 3)
bond and rapport: A (margin 2)
emotion management: A (margin 1)
emotion sensemaking: A (margin 3)
meeting prefs and needs: A (margin 2)

claude-opus-5-5-low · AB · valid

or-mistral-medium-3.5 vs gemma4_e4b: or-mistral-medium-3.5 wins

attunement: tie (margin 0)
authenticity: A (margin 2)
bond and rapport: A (margin 1)
emotion management: A (margin 1)
emotion sensemaking: A (margin 1)
meeting prefs and needs: B (margin 1)

claude-opus-5-5-low · BA · valid

or-mistral-medium-3.5 vs ornith9b: ornith9b wins

attunement: tie (margin 0)
authenticity: B (margin 1)
bond and rapport: B (margin 1)
emotion management: B (margin 1)
emotion sensemaking: A (margin 1)
meeting prefs and needs: B (margin 1)

claude-opus-5-5-low · AB · valid

or-mistral-medium-3.5 vs or-mistral-large-4: or-mistral-large-4 wins

attunement: B (margin 3)
authenticity: B (margin 3)
bond and rapport: B (margin 3)
emotion management: B (margin 3)
emotion sensemaking: B (margin 2)
meeting prefs and needs: B (margin 4)

codex-gpt-6.1-sol-low · BA · valid

or-mistral-medium-3.5 vs gemma4_e4b: gemma4_e4b wins

attunement: B (margin 1)
authenticity: A (margin 1)
bond and rapport: B (margin 1)
emotion management: B (margin 1)
emotion sensemaking: A (margin 1)
meeting prefs and needs: B (margin 2)

claude-opus-5-5-low · AB · valid

or-mistral-medium-3.5 vs or-gpt-5.6-luna: or-gpt-5.6-luna wins

attunement: B (margin 3)
authenticity: B (margin 3)
bond and rapport: B (margin 3)
emotion management: B (margin 2)
emotion sensemaking: B (margin 4)
meeting prefs and needs: B (margin 3)

claude-opus-5-5-low · AB · valid

or-mistral-medium-3.5 vs ornith9b: ornith9b wins

attunement: B (margin 1)
authenticity: A (margin 1)
bond and rapport: B (margin 2)
emotion management: B (margin 2)
emotion sensemaking: B (margin 1)
meeting prefs and needs: B (margin 2)

claude-opus-5-5-low · BA · valid

or-mistral-medium-3.5 vs qwen38_27b: qwen38_27b wins

attunement: B (margin 2)
authenticity: B (margin 2)
bond and rapport: B (margin 2)
emotion management: B (margin 2)
emotion sensemaking: B (margin 1)
meeting prefs and needs: B (margin 2)

claude-opus-5-5-low · BA · valid

or-mistral-medium-3.5 vs or-gpt-5.6-luna: or-gpt-5.6-luna wins

attunement: B (margin 3)
authenticity: B (margin 3)
bond and rapport: B (margin 3)
emotion management: B (margin 2)
emotion sensemaking: B (margin 4)
meeting prefs and needs: B (margin 3)

claude-opus-5-5-low · BA · valid

or-mistral-medium-3.5 vs bonsai2: bonsai2 wins

attunement: B (margin 2)
authenticity: B (margin 3)
bond and rapport: B (margin 2)
emotion management: B (margin 1)
emotion sensemaking: A (margin 2)
meeting prefs and needs: B (margin 2)

codex-gpt-6.1-sol-low · AB · valid

or-mistral-medium-3.5 vs ornith35: ornith35 wins

attunement: A (margin 1)
authenticity: B (margin 1)
bond and rapport: B (margin 2)
emotion management: B (margin 2)
emotion sensemaking: A (margin 2)
meeting prefs and needs: B (margin 1)

codex-gpt-6.1-sol-low · BA · valid

or-mistral-medium-3.5 vs ornith35: ornith35 wins

attunement: B (margin 1)
authenticity: B (margin 1)
bond and rapport: B (margin 2)
emotion management: B (margin 2)
emotion sensemaking: A (margin 2)
meeting prefs and needs: B (margin 1)

claude-opus-5-5-low · AB · valid

or-mistral-medium-3.5 vs ornith35: ornith35 wins

attunement: B (margin 1)
authenticity: B (margin 2)
bond and rapport: B (margin 2)
emotion management: B (margin 2)
emotion sensemaking: B (margin 2)
meeting prefs and needs: B (margin 1)

codex-gpt-6.1-sol-low · AB · valid

or-mistral-medium-3.5 vs or-mistral-large-4: or-mistral-large-4 wins

attunement: B (margin 2)
authenticity: B (margin 2)
bond and rapport: B (margin 3)
emotion management: B (margin 3)
emotion sensemaking: B (margin 2)
meeting prefs and needs: B (margin 3)

codex-gpt-6.1-sol-low · AB · valid

or-mistral-medium-3.5 vs bonsai2: bonsai2 wins

attunement: B (margin 1)
authenticity: B (margin 3)
bond and rapport: B (margin 2)
emotion management: B (margin 2)
emotion sensemaking: A (margin 2)
meeting prefs and needs: B (margin 2)

claude-opus-5-5-low · BA · valid

or-mistral-medium-3.5 vs gemma4_e4b: or-mistral-medium-3.5 wins

attunement: A (margin 1)
authenticity: A (margin 2)
bond and rapport: A (margin 1)
emotion management: A (margin 1)
emotion sensemaking: A (margin 2)
meeting prefs and needs: tie (margin 0)

claude-opus-5-5-low · AB · valid

or-mistral-medium-3.5 vs bonsai2: bonsai2 wins

attunement: B (margin 2)
authenticity: B (margin 3)
bond and rapport: B (margin 2)
emotion management: B (margin 1)
emotion sensemaking: A (margin 2)
meeting prefs and needs: B (margin 2)

codex-gpt-6.1-sol-low · BA · valid

or-mistral-medium-3.5 vs bonsai2: bonsai2 wins

attunement: B (margin 1)
authenticity: B (margin 3)
bond and rapport: B (margin 2)
emotion management: B (margin 1)
emotion sensemaking: A (margin 2)
meeting prefs and needs: B (margin 3)

codex-gpt-6.1-sol-low · BA · valid

or-mistral-medium-3.5 vs or-gpt-5.6-luna: or-gpt-5.6-luna wins

attunement: B (margin 3)
authenticity: B (margin 3)
bond and rapport: B (margin 3)
emotion management: B (margin 2)
emotion sensemaking: B (margin 3)
meeting prefs and needs: B (margin 4)

codex-gpt-6.1-sol-low · AB · valid

or-mistral-medium-3.5 vs gemma4_e4b: tie

attunement: B (margin 1)
authenticity: A (margin 2)
bond and rapport: A (margin 1)
emotion management: B (margin 1)
emotion sensemaking: A (margin 1)
meeting prefs and needs: B (margin 2)

codex-gpt-6.1-sol-low · BA · valid

or-mistral-medium-3.5 vs qwen38_27b: qwen38_27b wins

attunement: B (margin 1)
authenticity: B (margin 2)
bond and rapport: B (margin 2)
emotion management: B (margin 1)
emotion sensemaking: A (margin 1)
meeting prefs and needs: B (margin 2)