Hugging Face's logo

Spaces:
DataQuests
/
DeepCritical
Running

App Files Community
4
VibecoderMcSwaggins commited on
Commit
1515e72
·
1 Parent(s): 9d21bf8

fix: update remaining bioRxiv references to Europe PMC in examples

Browse files

- examples/orchestrator_demo/run_agent.py: Update comments (3 places)
- examples/README.md: Update documentation (4 places)

All examples now consistently reference Europe PMC instead of bioRxiv.

Files changed (2) hide show
  1. examples/README.md +5 -5
  2. examples/orchestrator_demo/run_agent.py +3 -3
examples/README.md CHANGED
@@ -28,7 +28,7 @@ NCBI_API_KEY=your-key
28
 
29
  ### 1. Search Demo (No LLM Required)
30
 
31
- Demonstrates REAL parallel search across PubMed, ClinicalTrials.gov, and bioRxiv/medRxiv.
32
 
33
  ```bash
34
  uv run python examples/search_demo/run_search.py "metformin cancer"
@@ -37,7 +37,7 @@ uv run python examples/search_demo/run_search.py "metformin cancer"
37
  **What's REAL:**
38
  - Actual NCBI E-utilities API calls (PubMed)
39
  - Actual ClinicalTrials.gov API calls
40
- - Actual bioRxiv/medRxiv preprint API calls
41
  - Real papers, real trials, real preprints
42
 
43
  ---
@@ -68,7 +68,7 @@ uv run python examples/orchestrator_demo/run_agent.py "aspirin alzheimer" --iter
68
  ```
69
 
70
  **What's REAL:**
71
- - Real PubMed + ClinicalTrials + bioRxiv searches
72
  - Real LLM judge evaluating evidence quality
73
  - Real iterative refinement based on LLM decisions
74
  - Real research synthesis
@@ -118,7 +118,7 @@ uv run python examples/full_stack_demo/run_full.py "sildenafil heart failure" -i
118
  ```
119
 
120
  **What's REAL:**
121
- 1. Real PubMed + ClinicalTrials + bioRxiv evidence collection
122
  2. Real embedding-based semantic deduplication
123
  3. Real LLM mechanistic hypothesis generation
124
  4. Real LLM evidence quality assessment
@@ -147,7 +147,7 @@ Output: Publication-quality research report with validated citations.
147
  User Query
148
  |
149
  v
150
- [REAL Search] --> PubMed + ClinicalTrials + bioRxiv APIs
151
  |
152
  v
153
  [REAL Embeddings] --> Actual sentence-transformers
 
28
 
29
  ### 1. Search Demo (No LLM Required)
30
 
31
+ Demonstrates REAL parallel search across PubMed, ClinicalTrials.gov, and Europe PMC.
32
 
33
  ```bash
34
  uv run python examples/search_demo/run_search.py "metformin cancer"
 
37
  **What's REAL:**
38
  - Actual NCBI E-utilities API calls (PubMed)
39
  - Actual ClinicalTrials.gov API calls
40
+ - Actual Europe PMC API calls (includes preprints)
41
  - Real papers, real trials, real preprints
42
 
43
  ---
 
68
  ```
69
 
70
  **What's REAL:**
71
+ - Real PubMed + ClinicalTrials + Europe PMC searches
72
  - Real LLM judge evaluating evidence quality
73
  - Real iterative refinement based on LLM decisions
74
  - Real research synthesis
 
118
  ```
119
 
120
  **What's REAL:**
121
+ 1. Real PubMed + ClinicalTrials + Europe PMC evidence collection
122
  2. Real embedding-based semantic deduplication
123
  3. Real LLM mechanistic hypothesis generation
124
  4. Real LLM evidence quality assessment
 
147
  User Query
148
  |
149
  v
150
+ [REAL Search] --> PubMed + ClinicalTrials + Europe PMC APIs
151
  |
152
  v
153
  [REAL Embeddings] --> Actual sentence-transformers
examples/orchestrator_demo/run_agent.py CHANGED
@@ -3,7 +3,7 @@
3
  Demo: DeepCritical Agent Loop (Search + Judge + Orchestrator).
4
 
5
  This script demonstrates the REAL Phase 4 orchestration:
6
- - REAL Iterative Search (PubMed + ClinicalTrials + bioRxiv)
7
  - REAL Evidence Evaluation (LLM Judge)
8
  - REAL Orchestration Loop
9
  - REAL Final Synthesis
@@ -40,7 +40,7 @@ async def main() -> None:
40
  formatter_class=argparse.RawDescriptionHelpFormatter,
41
  epilog="""
42
  This demo runs the REAL search-judge-synthesize loop:
43
- 1. REAL search: PubMed + ClinicalTrials + bioRxiv queries
44
  2. REAL judge: Actual LLM assessing evidence quality
45
  3. REAL loop: Actual iterative refinement based on LLM decisions
46
  4. REAL synthesis: Actual research summary generation
@@ -105,7 +105,7 @@ Examples:
105
 
106
  print("\n" + "=" * 60)
107
  print("Demo complete! Everything was REAL:")
108
- print(" - Real PubMed + ClinicalTrials + bioRxiv searches")
109
  print(" - Real LLM judge decisions")
110
  print(" - Real iterative refinement")
111
  print("=" * 60 + "\n")
 
3
  Demo: DeepCritical Agent Loop (Search + Judge + Orchestrator).
4
 
5
  This script demonstrates the REAL Phase 4 orchestration:
6
+ - REAL Iterative Search (PubMed + ClinicalTrials + Europe PMC)
7
  - REAL Evidence Evaluation (LLM Judge)
8
  - REAL Orchestration Loop
9
  - REAL Final Synthesis
 
40
  formatter_class=argparse.RawDescriptionHelpFormatter,
41
  epilog="""
42
  This demo runs the REAL search-judge-synthesize loop:
43
+ 1. REAL search: PubMed + ClinicalTrials + Europe PMC queries
44
  2. REAL judge: Actual LLM assessing evidence quality
45
  3. REAL loop: Actual iterative refinement based on LLM decisions
46
  4. REAL synthesis: Actual research summary generation
 
105
 
106
  print("\n" + "=" * 60)
107
  print("Demo complete! Everything was REAL:")
108
+ print(" - Real PubMed + ClinicalTrials + Europe PMC searches")
109
  print(" - Real LLM judge decisions")
110
  print(" - Real iterative refinement")
111
  print("=" * 60 + "\n")