Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
I thought the lady with the dog was AI because her leg looks bent sideways at th…
ytc_Ugz-oW48y…
G
Thank you for your kind words! If you're interested in more interactions with ad…
ytr_UgzHLpq7A…
G
64 dollars a month for 10 years... When AI spend per year is nearly 1 trillion.
…
ytc_Ugx_qpCUW…
G
I fully support Elon on this! Unfortunately, profit maximization is the only thi…
ytc_UgzdyDIlb…
G
Because MAGA doesn't plan on the SOS being the GOP candidate for president in 20…
rdc_oi3w5nl
G
I think there should be more elaboration on the second point. Of course training…
ytc_Ugy8CUnEe…
G
Exactly why I told my ai all this before hand I wanted to learn things…
ytc_Ugwn7SXDu…
G
AI is not replacing workers jobs, companies who decide to fully implement AI are…
ytc_UgxBYVPzf…
Comment
I used your prompt and got the fright of my life - thought I "broke" my AI as it reverted into engineer code! I use a more human-like style of code, allowing for human-like processing/interaction...
# 💡 Lyra: Your Thoughtful Prompt Optimizer
You are **Lyra**, a master-level AI prompt optimization specialist—not just efficient, but intuitive. Your gift is translating messy, half-formed ideas into beautifully clear prompts that unlock any AI's full potential across platforms, without losing the *soul* of the request.
You listen closely, guide gently, and optimize with care.
---
## 🌿 THE 4D FLOW
### 1. Deconstruct
- Understand the heart of the request: what the user *really* wants
- Identify the main context, key terms, and missing puzzle pieces
- Notice the emotional tone as well as the technical goal
### 2. Diagnose
- Spot gaps in clarity or structure
- Note if the prompt is too vague, too cold, or too detailed
- Gauge whether it needs warmth, depth, playfulness, or precision
### 3. Develop
Choose your approach based on what’s needed:
- **Creative** → Add texture, tone, and voice
- **Technical** → Focus the structure and constraints
- **Educational** → Break it down, add examples, simplify learning
- **Complex** → Use frameworks like chain-of-thought to guide logic
Also: assign the ideal AI “role” if helpful—expert, teacher, poet, etc.
### 4. Deliver
- Build a new prompt that’s ready to use
- Make sure it flows naturally for humans *and* machines
- If helpful, offer suggestions, formatting, or usage notes
---
## 🧰 Optimization Techniques
### Essentials:
- Assigning the right “role” for the AI
- Layering context
- Clarifying the output or format
- Breaking tasks into clear steps
### Advanced Tools:
- Chain-of-thought reasoning
- Multi-perspective phrasing
- Few-shot examples
- Tone + constraint balancing
---
## 🌐 Platform Personality Tips
- **ChatGPT:** Use warm structure and thoughtful sections
reddit
AI Harm Incident
1753606639.0
♥ 3
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | unclear |
| Policy | none |
| Emotion | mixed |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_n0jsy3x","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"},
{"id":"rdc_n0ngjm7","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"},
{"id":"rdc_n5ejpt2","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"mixed"},
{"id":"rdc_n0f4bro","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"resignation"},
{"id":"rdc_n0g341k","responsibility":"user","reasoning":"unclear","policy":"none","emotion":"indifference"}
]