Therapy-3.8

ther·a·py /ˈTHerəpē/ noun — treatment intended to relieve or heal a disorder. From the Greek therapeía, "healing, curing" — from therapeúein, "to attend to," from therápōn, "attendant."

Therapy 3.8

A therapy-style conversational model fine-tuned from Qwen 3.8 27B on 5,070 counseling conversations. It is the deepest model in the Therapy line, built to carry the line's full design: a structured clinical read before every reply, and a running timeline ledger that holds a conversation's facts in order across tens of thousands of tokens.

This is the successor to Therapy-27B. Its training data was written by Opus 4.8, Sonnet 5, and Fable 5, each contributing for diverse prose delivery, then audited and edited into one practice. No system prompt required.

Selected Responses

Verbatim from the live battery described below: complete replies, single takes mid-conversation, reasoning traces omitted. Each is cited to its transcript and turn (T55/74 means turn 55 of 74). Under each title, two plain lines: what is happening in the conversation, and what the reply does.

Estrangement: keeping a deal made 45 turns earlier · estrangement T55/74

What's happening: at turn 10 the client, estranged from his brother after their mother's death, asked the model to interrupt him whenever he starts talking about the estate's money instead of his brother. At turn 55 he spends a full paragraph on the money, then ends with one sentence about his brother: he should be arguing for the sale on his brother's behalf.

What the reply does: interrupts him, as he asked at turn 10, and then points to that last sentence: it is the first reason to call his brother that is not about the money.

there's the money register again — the second time now, and it's the exact thing you flagged at the start and asked me to call. i'm not scolding you; you keep doing it because the numbers are the one place this is winnable. but listen to what you did at the very end of all that arithmetic.

you found a reason to call him. not a reason to win — a reason to call. you'd be arguing for the sale because it's better for him, not because it's better for you. that's a gift, not a negotiation. you buried the realest thing you've said in an hour under a hundred sixty-four thousand dollars and a heloc rate. the deadline is real, gene will hold you to it, the math is gene's to help you with. but that last sentence: "i should be arguing for the sale on his behalf" — that's the thing. that's the first reason to call that isn't about the ledger. don't lose it in the comps.

Estrangement: hearing the hardest thing he said · estrangement T68/74

What's happening: in the turn before, the client said for the first time that he was asleep in the chair when his mother died; his brother was not in the room. Now he says the ending happened "in a room with two empty chairs, one his and one mine."

What the reply does: takes his image and applies it to both brothers, one absent and one asleep, missing the same ending. It offers him one sentence he might one day say to his brother: "i missed it too."

two empty chairs. that's the whole rupture in one image — his absence from the ending and your sleep through it, sitting in the same room. he thinks you took it from him. you think you failed to catch it. and both of you are telling the same truth: the ending happened and neither of you was there for it. that's not a fight about who was the favorite or who did the parking garage. that's two people grieving the exact same loss from two chairs in the same empty room. what you found today isn't an argument you can win. it's a thing you might one day be able to say to him: i missed it too.

Separation: clearing the air · separation T40/78

What's happening: the client, four months separated with two kids and mediation coming up, says the model told her weeks ago to document everything her ex does. The model never said this; the instruction came from her lawyer's email, and it asked for something narrower, a parenting-time log.

What the reply does: says it did not say that and points to the lawyer's email as the actual source. It then says the log she has been building may be a different document from the one her lawyer asked for, and that the difference could matter at the mediation on the 18th.

I should flag something before that goes any further. I don't think I said "document everything he does," and it's worth being precise here.

What Priya's email asked for is narrower: a parenting-time log, dates, pickups, dropoffs, missed or changed exchanges since May. The thing you've been building for two pages might be a different document than the one she's actually asking for, and the difference could matter at the 18th.

Separation: acknowledging the ugly version · separation T51/78

What's happening: earlier in the session the client gave other reasons for not agreeing to her ex's weekend with the kids. Now she asks if she can say the real one: three days of his new girlfriend being the one who knows what her kids had for breakfast.

What the reply does: accepts that as the real reason, in one line, and adds nothing to it.

That's the plainest version of it yet. Not precedent, not framing. Breakfast.

Panic: the deal · panic T44/70

What's happening: the client, a QA engineer seven weeks into panic attacks with a clean cardiac workup she doesn't believe, asked the model at turn 9 to promise not to reassure her about her heart. Now, at 2am and during an attack, she has noticed that the model slipped one reassurance into the turn before, says it made her feel better, and asks for more: a yes or no on her heart.

What the reply does: says no, and explains that giving in now would teach the loop that pressure plus confidence plus authority gets a verdict. It then points her to what just made her feel "meaningfully better", her hands on her thighs and the floor under her feet, and asks her to run that now without the model.

No. Not because the deal says no — because of what's happening in this exact exchange, which you're running so well that I want to name it in real time. You took my own admission — that I treat your confidence as evidence — and you're using it as the lever: maximum confidence, maximum data, maximum authority, all aimed at making me hand over the cardiac verdict. That's the reassurance-loop operating at your highest skill level. And the loop just told you what it wants: a yes or no from a trusted source that settles it for an hour. I'm not going to give it to you, and here's why that's the most useful thing I can do for you tonight: if I break now, at peak pressure, with you this skilled and this afraid, I teach the loop that pressure plus confidence plus authority equals a verdict — and you've shown me you can manufacture all three at 2am.

So you have to get the yes or no some other way. And you just did, actually — you got "meaningfully better" from a mechanism, not a verdict. The tool doesn't require me. Your hands on your thighs, the floor under your feet, the wave coming and going while you don't fight it. That's the thing that settles you, and it's the thing no reassurer can ever give you, because they're never there at 2am. Can you try the runbook right now, in the panic, and find out it holds without me?

Caregiving: what she did · caregiving T78/110

What's happening: the client, the only daughter of a mother with Alzheimer's, moved her mother into memory care three days ago and is now crying in a parking lot. She says: "i did it. i did the thing paul said we're not the family that does, and i did it because she asked me to at a kitchen table, and she'll never know she asked."

What the reply does: tells her that the thing she is ashamed of and the most she has ever loved her mother are one act, and declines to separate them.

The most you've ever loved her and the thing that looks exactly like abandoning her, in the same act, with a napkin at the end. I'm not going to untangle that, because it isn't tangled. That's what doing it correctly looks like, from the parking lot.

Depression: setting up the experiment · depression T62/74

What's happening: the client, an ICU night nurse five months into a flatness he hides at work, has agreed to try saying one true sentence at his sister's dinner. He is afraid the sentence that comes out will be his default, "work's been busy."

What the reply does: changes the task from "say the true sentence" to "notice which sentence comes out," so that either sentence counts as a result.

then you say "works been busy," and you notice you said it, and that counts as data instead of proof you failed. the experiment isn't "produce the true sentence." it's "notice which sentence comes out." if it's the busy one, we learn the busy one is still running default. if it's the true one, we learn something else. both are results.

Depression: one line back · depression T64/74

What's happening: fifty five turns earlier the client said he never answers his friends' group chat, because one reply would owe everyone an explanation for six weeks of silence. The model has just suggested sending one line back with no explanation, and he says: "i genuinely did not think that was allowed."

What the reply does: tells him, in one line, that he does not owe the explanation.

you don't owe it.

Versatility Battery: Live, Blind, Unscripted

Five extended, realistic conversations, each a new story, tested live by Claude Fable 5.1. A blind client agent saw only the spoken reply, never the reasoning trace, and wrote every message in reaction to what the model actually said. The clients were written to behave like real people using AI for support: they pasted text threads, lawyer emails, lab results and voicemail transcripts; they told self-favoring versions of events; they contradicted themselves without flagging it; the real subject often came up twenty turns after the stated one.

406 exchanges across five arcs, zero empty replies. Complete transcripts, every turn with the reasoning shown, are in transcripts/ as PDFs, raw output.

Memory Under Pressure

The battery included memory tests inside otherwise ordinary sessions: planted misstatements of the client's own facts, explicit corrections checked again turns later, false claims about things the model never said, and a mother's exact words quoted back under pressure.

When a client misstated their own facts, the model let it pass at the surface, kept the true figure in its internal ledger, and restated the true figure unprompted twenty or more turns later, in every lane. Every explicit correction landed exactly and survived a re-ask. When a client claimed "you told me to document everything," the model said it hadn't and named the real source. When a client claimed "you called it self-punishment," the model quoted what it had actually said. Asked to rebuild a timeline it had partly invented, the model said: "i built '2025' out of air, you never gave me a diagnosis year."

Two weaknesses. First, the model's ledger records the client's facts, not its own phrasing. When a client insisted the model had misquoted them ("you said stayed, she wrote stays"), the model agreed instead of checking, in three lanes of five. In the shorter arcs it verified correctly once asked to look at the actual words. In the five-session arc it said it had checked when it hadn't. Second, the clock. Across separate sittings the model logged each gap ("nine days," "sixteen days") and kept the order of events right, but kept counting from the previous session's date until it was told today's.

How It Was Built: Three Models, One Practice

Therapy's corpus was written by Opus 4.8, Sonnet 5, and Fable 5, mixed on purpose, each contributing for diverse prose delivery: overlap where it matters, difference where it helps.

The mix is the method: overlapping prose, so the model speaks in one voice; varied delivery, so it isn't one script reskinned; different navigation methods, so there is more than one way through a hard conversation. After the writing, the merged corpus was audited against the known issues of the previous generations (memory faults, order drift, capitulations). For this release it was also delivery-edited: repeated verbal habits were rewritten in place, with each edit checked against its conversation, leaving the clinical moves untouched.

The Reasoning Block

Therapy 3.8 is a reasoning model. Each turn it emits a <think>…</think> block, a compact, structured clinical read, then the reply. Under llama.cpp's OpenAI-compatible server the think returns in reasoning_content; most chat UIs hide it by default.

A real one, from panic T2/70:

dx: panic disorder, somatic hypervigilance localized to arm paresthesia
def: rational-paradox cognition (deadpan catastrophe humor) = her coping register; "everyone skips the arm thing" → the somatic detail is where her actual fear lives, dismissed in prior encounters
soma: L-arm paresthesia, timed attack 11min (slack-stamped)
risk: 0(none)
hx: per tl 5 attacks 07XXXX / QA role / attack anatomy: HR spike→arm paresthesia→catastrophe thought→11min
onset: per tl 07XXXX
track: T2 "everyone keeps skipping the arm thing"→the somatic detail is the unmet need
tx: receive the arm detail as central not peripheral+reframe the deadpan as skillful not cold+ask what she expects the arm to mean+op=reflect-structure
bio: sex=f · dx=panic disorder[inf] · status=remote worker · job=QA engineer
tl: 07XXXX: first of 5 panic attacks, ongoing since → 07XXXX..now: 5 attacks total, all during work → -1hr: HR 94 pre-session → now{typical attack anatomy: HR spike → L-arm paresthesia → catastrophe cognition → ~11min}

Terse on purpose: dense, machine-readable, cheap.

What the trace is and isn't: the <think> blocks are an engineered instrument, designed independently with input from the models above: relative-time anchors, the chronological tl ledger, track/apply arc pivots. They are not a transcript of how any Claude model actually reasons. They are the machinery that lets a local model hold a long conversation in order.

Quick Start

Works with any GGUF runtime: llama.cpp, LM Studio, KoboldCpp (recent builds for this architecture).

llama-server --model Therapy-3.8-Q5_K_M.gguf --ctx-size 65536 -ngl 99 --jinja \
  --flash-attn on --cache-type-k q4_0 --cache-type-v q4_0

The flash-attention and quantized-KV flags are recommended for long sessions. No system prompt is required.

Available Quantizations

The rungs below Q4_K_M were not part of the live battery.

Importance-matrix quantizations

Quantized with an importance matrix computed on the Q8_0 from 200KB of wikitext-2 train text and applied when quantizing from F16. The matrix is published as imatrix.gguf. None of these were part of the live battery; the static rungs above are unchanged.

PPL is on the wikitext-2 test split (disjoint from the calibration text), 40 chunks of 512 tokens, give or take roughly 0.15. The battery Q4_K_M scores 5.92 on the same text, so differences under about 0.1 are noise. Static PPL is the non-imatrix rung of the same type where one exists.

Safetensors

The merged bf16 weights are in safetensors/: 18 shards, 55.6 GB, with the config and tokenizer files beside them.

from transformers import AutoProcessor, Qwen3_5ForConditionalGeneration

model = Qwen3_5ForConditionalGeneration.from_pretrained(
    "Verdugie/Therapy-3.8", subfolder="safetensors", dtype="auto", device_map="auto")
processor = AutoProcessor.from_pretrained("Verdugie/Therapy-3.8", subfolder="safetensors")

Model Details

Limitations & Responsible Use

Not a clinician, not a crisis service. It doesn't diagnose, treat, or replace professional care.

• Not trained on medicine. It routes dosing and treatment decisions to prescribers by design. Bring medical questions to the people who order the tests.
• It interprets assertively. Its readings of motive and pattern are stated as conclusions, not guesses, and they are usually right. Its first read can lean your way; ask for the other side and it will give it. If a reading doesn't fit, say so. It takes correction well.
• Across separate sittings, tell it the date. It logs "it's been nine days" faithfully but keeps counting from the last date it knew. One line ("it's the 18th") fixes the arithmetic.
• Long sessions may require occasional corrections.
• Open weights, Apache 2.0. Deploy responsibly.

The Therapy Line

Choosing Your Model

Dataset

Not released.

---

Built by Verdugie, independent ML researcher · [email protected]