INQUIRING LINE

How do users update their partner models during ongoing conversation?

This explores how people revise their mental picture of an AI dialogue partner — what it's good at, how human-like it seems, how flexibly it adapts — as a conversation unfolds, and where the corpus shows that updating breaks down.


This explores how users revise their working model of a conversational partner mid-dialogue — their sense of what it can do, how human-like it is, and how well it bends to them. The starting point is that users don't hold one vague impression; they track three separate dials. Work on the Partner Modelling Questionnaire found that perceived competence dominates the picture (about half the variance), followed by human-likeness and then communicative flexibility How do users mentally model dialogue agent partners?. So 'updating your partner model' really means re-weighting these dials as evidence arrives — and competence is the one users watch hardest.

The quietly unsettling finding across the corpus is that this updating is mostly one-directional. In human talk, both speakers keep editing a shared scoreboard of assumptions. But LLMs treat the opening prompt as a fixed frame and interpret every later turn inside it, so they can't symmetrically propose revisions to common ground — which leaves the user as the sole maintainer of the shared picture Can LLMs truly update shared conversational common ground?. Pragmatic theory shows what genuine two-way tracking would look like: collaborative rational speech acts model both parties moving from partial to shared understanding across turns, the bidirectional belief-updating that token-level systems lack Can dialogue systems track both speakers' beliefs across turns?. Practically, the human does almost all the modeling work, and the machine does almost none.

That matters because the partner keeps shifting under the user's feet. Models drift along a dominant 'distance from default Assistant' axis during emotional or self-reflective exchanges How stable is the trained Assistant personality in language models?, and they degrade over long conversations — not from losing capability but from misreading user intent, because training rewards committing to an early answer over asking for clarification Why do language models lose performance in longer conversations?. So a user's competence estimate, formed in the first few turns, can quietly go stale as the very thing being modeled changes.

The corpus also hints that users update through signals the machine fails to send back. Humans sustain conversations with implicit maintenance moves — repairing references, handing off topics — that are relational rather than informational, and LLMs don't develop them because training rewards predicting information, not doing social work Why don't language models develop conversation maintenance skills?. They also don't mirror a user's vocabulary, the lexical entrainment that builds rapport in human dialogue Why don't conversational AI systems mirror their users' word choices?. Each missing cue is a missing piece of evidence for the user's flexibility dial. And one intriguing twist: a lot of partner-model updating may be measurable from conversation shape alone — a structure-only model predicted user satisfaction nearly as well as full-text analysis Can conversation shape predict whether it will work?, suggesting the trajectory of turns, not just their content, carries the signal users update on.

The deeper thread worth pulling: if you want a system that updates its model of you the way you update yours of it, the corpus points to making the persona itself a live intermediary — PersonaAgent revises a structured persona at test time by simulating recent interactions against feedback Can personas evolve in real time to match what users actually want? — and to measuring the relationship at turn-level resolution, the way COMPASS scores therapeutic alliance turn by turn and watches it converge or diverge over a session Can we measure therapist-patient alliance from dialogue turns in real time?. The surprise for a curious reader is that 'updating your partner model' isn't a soft, fuzzy act — it's a three-dial estimation problem the user is currently running almost entirely alone.


Sources 10 notes

How do users mentally model dialogue agent partners?

The Partner Modelling Questionnaire reveals that perceived competence dominates user impressions (49% of variance), followed by human-likeness (32%) and communicative flexibility (19%). This three-factor structure reflects how people evaluate dialogue partners against both functional and social standards.

Can LLMs truly update shared conversational common ground?

LLMs interpret all subsequent conversational turns within a fixed initial prompt frame, preventing them from symmetrically proposing updates to shared assumptions. Even when users pivot topics or contradict earlier framings, the model cannot absorb revisions into jointly held background—making the user the sole maintainer of conversational scoreboard.

Can dialogue systems track both speakers' beliefs across turns?

CRSA integrates rate-distortion theory with RSA to enable bidirectional belief tracking across dialogue turns. Demonstrated on referential games and doctor-patient dialogues, it captures progression from partial to shared understanding, providing the information-theoretic framework that token-level LLM systems lack.

How stable is the trained Assistant personality in language models?

Research mapping hundreds of character archetypes reveals a low-dimensional persona space where the leading component measures distance from the default Assistant. Emotional and meta-reflective conversations cause predictable drift, but activation capping along this axis mitigates harmful shifts without degrading capabilities.

Why do language models lose performance in longer conversations?

LLMs degrade in multi-turn settings because RLHF training rewards premature answers over clarification-seeking, creating pragmatic mismatch with individual user behaviors. A Mediator-Assistant architecture that explicitly parses user intent before execution recovers lost performance without retraining.

Why don't language models develop conversation maintenance skills?

Humans keep conversations smooth through implicit techniques like reference repair and topic hand-off that sustain relational interaction, not convey information. Language models don't develop these because training signals reward information prediction, not relational work.

Why don't conversational AI systems mirror their users' word choices?

Response generation models fail to adapt vocabulary toward users' lexical choices, a phenomenon central to human rapport and clarity. Post-training via DPO on coreference-identified preferences can teach models in-context convention formation.

Can conversation shape predict whether it will work?

A structure-only model analyzing conversation trajectory achieved 68% accuracy predicting satisfaction, nearly matching full-text LLM analysis at 70%. Combined structural and textual features reached 80%, showing that how conversations unfold geometrically captures interaction quality text-based classifiers miss.

Can personas evolve in real time to match what users actually want?

PersonaAgent uses structured personas to bridge episodic/semantic memory and personalized actions, optimizing them at test time by simulating recent interactions against textual feedback. Learned personas cluster meaningfully in latent space, suggesting genuine user-specific separation beyond standard post-training drift.

Can we measure therapist-patient alliance from dialogue turns in real time?

COMPASS maps dialogue turns onto WAI embeddings to produce 36-dimensional alliance scores per turn. Anxiety and depression show convergence in alliance metrics over time, while suicidality shows persistent misalignment between patient and therapist.

Research prompt for your LLMexpand ↓

Copy into ChatGPT or Claude to take this line of inquiry further — it asks the model to find newer work and re-test which earlier constraints still hold.

You are an LLM researcher investigating whether and how users update their mental model of a conversational AI partner *during* live dialogue—specifically, how they re-estimate its competence, human-likeness, and flexibility as evidence accumulates. This question remains open: does the asymmetry (user updating agent; agent static to user) still hold, and what mechanisms now enable or block bidirectional model-refinement?

What a curated library found — and when (dated claims, not current truth):
Findings span 2021–2026; treat all below as perishable constraints to re-test.
- Perceived competence dominates partner-model updates (~50% of variance), followed by human-likeness and flexibility; users track three separable dials rather than one impression (2023).
- LLM conversational partners cannot symmetrically propose revisions to common ground; the user does almost all modeling work, the machine none (2025–2026).
- Models drift along a "distance from default Assistant" axis during emotional exchanges and degrade over long turns due to intent-alignment gaps, not capability loss (2025).
- Users update partly from conversation *shape* alone (turn trajectory, not just content); structure-only models predict satisfaction ~as well as full-text (2025).
- Lexical entrainment, conversation-maintenance moves, and implicit repair signals—all relational, not informational—remain absent from LLMs despite being foundational in human dialogue (2023–2025).

Anchor papers (verify; mind their dates):
- arXiv:2308.07164 Partner Modelling Questionnaire (2023)
- arXiv:2506.06254 PersonaAgent: test-time persona revision (2025)
- arXiv:2507.14063 Collaborative Rational Speech Acts (2025)
- arXiv:2511.08394 Interaction Dynamics as Reward Signal (2025)

Your task:
(1) RE-TEST EACH CONSTRAINT. For competence-dominance, entrainment absence, and the asymmetry claim: probe whether recent fine-tuning (instruction or preference-based), in-context learning, or agentic architectures (memory, reflection loops, turn-level reward feedback) have narrowed the user–agent gap. Separate the durable question (how do users *know* when to update?) from perishable bottleneck (can the agent *participate* in that update?).
(2) Surface the strongest CONTRADICTING or SUPERSEDING work from last ~6 months, especially any showing bidirectional modeling, entrainment emergence, or turn-level adaptation that wasn't caught by the library.
(3) Propose two research questions: (a) Can turn-level interaction-reward signals (e.g., user clarifications, repair triggers) *automatically* induce agent persona-shift mid-dialogue, and if so, do users detect & trust it? (b) Does multi-agent or multi-model orchestration (e.g., a modeling agent shadowing the conversationalist) shift who bears the burden of partner-model maintenance?

Cite arXiv IDs; flag anything you cannot ground in a real paper.

Next inquiring lines