IPDA Debate Canonical Model v3
Collection
Essential models and datasets used to build the IPDA debate canonical model. Includes ORPO, GRPO iterations, SFT distillation, and golden samples. • 15 items • Updated
prompt stringlengths 6.47k 9.04k | response stringlengths 5.77k 13.2k | score float64 0.7 1 | source stringclasses 2
values |
|---|---|---|---|
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this 1AR situation carefully using the JAM framework.
## JUDGE Analysis
This judge works on CRISPR applications and bioethics - she's an expert in this field. She's "deeply conflicted" about germline editing, seeing both therapeutic promise and enhancement concerns. Key preferences:
- Don't stra... | 0.95 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this 2AR situation using the JAM framework.
## JUDGE Analysis
The judge teaches argumentation theory and did policy debate at Northwestern. She values:
- Pedagogy - "would a student learn from this?"
- Building understanding
- Making ideas accessible
- Genuine comprehension over winning tricks
-... | 0.95 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this 1AR situation using the JAM framework.
## JUDGE Analysis
- Retired librarian with no debate background
- Values clarity, everyday examples, and understanding
- Decides based on "who made the most sense"
- Will be lost by technical jargon or spreading
- The AC used Maria's story effectively ... | 0.95 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this NR (Negative Rebuttal) situation using the JAM framework.
## JUDGE Analysis
The judge is a retired librarian with no debate experience. Key implications:
- Needs CLARITY above all else
- Everyday examples will resonate
- Stories are powerful (they used Selena's story effectively)
- Technica... | 0.96 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this debate situation using the JAM framework.
## JUDGE Analysis
The judge works on mechanistic interpretability and is "optimistic but realistic about tractability at scale." Key points:
- Wants genuine understanding of what interpretability research has and hasn't shown
- Doesn't want conflati... | 0.95 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this debate using the JAM framework.
## JUDGE Analysis
The judge works on CRISPR applications and bioethics. She's interested in nuanced engagement with difficult cases, not strawmanning. Key observations:
- She values nuanced, scientifically grounded arguments
- She's deeply conflicted about co... | 0.95 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this debate situation using the JAM framework.
## JUDGE Analysis:
- Retired librarian, no debate background
- Wants clarity and understanding
- Uses everyday examples
- Decides based on who "made the most sense"
- The AC actually used a librarian example (Maria wanting to be a librarian) - this ... | 0.9875 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this debate situation using the JAM framework.
## JUDGE Analysis
Professor Jones teaches argumentation theory and did policy debate at Northwestern. Key preferences:
- Values pedagogy - "would a student learn from this?"
- Rewards building understanding and accessibility
- Values genuine compreh... | 0.95 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this debate situation using the JAM framework.
## JUDGE Analysis
Ms. The Judge is a 10th grade English teacher with no debate background. She values:
- Clear communication
- Logical structure
- Persuasive essay-style arguments
- No technical jargon
- Help following along
This means I should avo... | 0.95 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this 2AR situation using the JAM framework.
## JUDGE Analysis
- Retired librarian, no debate background
- Wants to UNDERSTAND - clarity is paramount
- Prefers everyday examples and human stories
- Decides based on "who made the most sense"
- For lay judges: stories > abstract principles
- Need t... | 0.95 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this 2AR situation using the JAM framework.
## JUDGE Analysis
The judge is a retired librarian with no debate background. Key preferences:
- Wants to understand what's happening
- Values clarity and everyday examples
- Decides based on "who made the most sense"
- Stories > abstract principles
- ... | 0.95 | opus |
[SYSTEM]
Your input fields are:
1. `topic` (str): The debate resolution
2. `speech_type` (str): Type of speech: AC, NC, 1AR, NR, 2AR
3. `side` (str): AFF or NEG
4. `debate_state` (str): Current debate state including prior speeches
5. `available_tactics` (str): Available tactics to choose from
6. `judge_bio` (str): Jud... | <think>
Let me analyze this 2AR situation using the JAM framework.
## JUDGE Analysis
The judge teaches argumentation theory and did policy debate at Northwestern. Key characteristics:
- Values pedagogy - "would a student learn from this?"
- Rewards building understanding and making ideas accessible
- Values genuine co... | 0.95 | opus |
No dataset card yet