File size: 16,974 Bytes
d7fb055 f6ae12d dc6e87c d7fb055 e4a5f6d dc6e87c d7fb055 dc6e87c d7fb055 013a35e d7fb055 013a35e d7fb055 013a35e d7fb055 852294d d7fb055 8425946 d7fb055 b9aa360 d7fb055 dc6e87c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 |
# Main application entry point for AI Co-Scientist
import os
import argparse
import logging
import json
import time
from datetime import datetime
from pathlib import Path
from typing import Dict, List, Any, Optional
from .agents.agent_factory import AgentFactory
from .tools.search_tool import SearchTool
from .tools.reasoning_tool import ReasoningTool
from .tools.citation_tool import CitationTool
from .utils.logger import setup_logging
from .utils.validators import validate_research_goal, validate_hypothesis
from .utils.exceptions import ConfigError, ValidationError
from .config.config import (
AGENT_DEFAULT_MODEL,
AGENT_DEFAULT_TEMPERATURE,
LOG_LEVEL,
LOG_TO_FILE,
MAX_ITERATIONS,
MAX_TOKENS
)
class AICoScientist:
"""Main class for the AI Co-Scientist system.
This class coordinates the multi-agent scientific research workflow,
managing the interactions between different specialized agents to generate
and refine scientific hypotheses.
"""
def __init__(self, config: Optional[Dict[str, Any]] = None):
"""Initialize the AI Co-Scientist system.
Args:
config: Optional configuration dictionary to override defaults
"""
# Initialize logger
self.logger = setup_logging(log_level=LOG_LEVEL, log_to_file=LOG_TO_FILE)
self.logger.info("Initializing AI Co-Scientist system")
# Load configuration
self.config = {
"model": AGENT_DEFAULT_MODEL,
"temperature": AGENT_DEFAULT_TEMPERATURE,
"max_iterations": MAX_ITERATIONS,
"max_tokens": MAX_TOKENS,
}
# Override with provided config
if config:
# Ensure model is never None
if config.get("model") is None:
config["model"] = AGENT_DEFAULT_MODEL
self.config.update(config)
self.logger.info(f"Configuration overridden with custom values")
self.logger.info(f"Using model: {self.config['model']}")
# Initialize agent factory
self.agent_factory = AgentFactory()
# Initialize tools
self.tools = {
"search": SearchTool(),
"reasoning": ReasoningTool(),
"citation": CitationTool()
}
# Create supervisor agent
self.supervisor = self.agent_factory.get_supervisor(
model=self.config["model"],
temperature=self.config["temperature"]
)
# Initialize research state
self.research_goal = None
self.hypotheses = []
self.ranked_hypotheses = []
self.final_report = None
def set_research_goal(self, goal: str) -> bool:
"""Set the research goal for the system.
Args:
goal: The research goal to pursue
Returns:
True if the goal was successfully set, False otherwise
Raises:
ValidationError: If the goal is invalid
"""
self.logger.info(f"Setting research goal: {goal}")
# Validate the research goal
is_valid, issues = validate_research_goal(goal)
if not is_valid:
error_msg = f"Invalid research goal: {', '.join(issues)}"
self.logger.error(error_msg)
raise ValidationError(error_msg)
self.research_goal = goal
self.logger.info(f"Research goal set successfully")
return True
def generate_hypotheses(self, count: int = 5) -> List[Dict[str, Any]]:
"""Generate scientific hypotheses based on the research goal.
Args:
count: Number of hypotheses to generate
Returns:
List of generated hypotheses with metadata
Raises:
ValueError: If research goal is not set
"""
if not self.research_goal:
raise ValueError("Research goal must be set before generating hypotheses")
self.logger.info(f"Generating {count} hypotheses for research goal: {self.research_goal}")
# Get generation agent
generation_agent = self.agent_factory.get_agent(
"generation",
model=self.config["model"],
temperature=self.config["temperature"] + 0.2 # Slightly higher temperature for creativity
)
# Generate hypotheses
hypotheses = generation_agent.generate_hypotheses(self.research_goal, count=count)
# Convert 'hypothesis' key to 'statement' for downstream compatibility
for h in hypotheses:
if "hypothesis" in h:
h["statement"] = h["hypothesis"]
# No validation or filtering: accept all generated areas of interest
for h in hypotheses:
area = h["statement"]
# Simple transformation: create 2-3 research questions per area
h["research_questions"] = [
f"What are the key mechanisms underlying: {area}?",
f"How could {area} be investigated in different populations or contexts?",
f"What are the potential implications of {area} for future research or applications?"
]
self.hypotheses = hypotheses
self.logger.info(f"Generated {len(hypotheses)} areas of interest")
return hypotheses
def rank_hypotheses(self) -> List[Dict[str, Any]]:
"""Rank the generated hypotheses by quality and relevance.
Returns:
List of ranked hypotheses with scores
Raises:
ValueError: If no hypotheses have been generated
"""
if not self.hypotheses:
raise ValueError("No hypotheses have been generated to rank")
self.logger.info(f"Ranking {len(self.hypotheses)} hypotheses")
# Get ranking agent
ranking_agent = self.agent_factory.get_agent("ranking")
# Rank hypotheses
ranked_hypotheses = ranking_agent.rank_hypotheses(
self.research_goal,
self.hypotheses
)
self.ranked_hypotheses = ranked_hypotheses
self.logger.info(f"Ranked {len(ranked_hypotheses)} hypotheses")
return ranked_hypotheses
def refine_hypotheses(self, iterations: int = 3) -> List[Dict[str, Any]]:
"""Refine the top hypotheses through multiple iterations.
Args:
iterations: Number of refinement iterations
Returns:
List of refined hypotheses
Raises:
ValueError: If no ranked hypotheses are available
"""
if not self.ranked_hypotheses:
raise ValueError("No ranked hypotheses are available for refinement")
# Get top 3 hypotheses for refinement
top_hypotheses = self.ranked_hypotheses[:3]
self.logger.info(f"Refining top {len(top_hypotheses)} hypotheses through {iterations} iterations")
# Get agents for refinement
reflection_agent = self.agent_factory.get_agent("reflection")
evolution_agent = self.agent_factory.get_agent("evolution")
proximity_agent = self.agent_factory.get_agent("proximity")
refined_hypotheses = top_hypotheses.copy()
for i in range(iterations):
self.logger.info(f"Refinement iteration {i+1}/{iterations}")
for j, hypothesis in enumerate(refined_hypotheses):
# Get feedback from reflection agent
feedback = reflection_agent.review_hypothesis(
hypothesis.get("hypothesis", hypothesis.get("statement")),
self.research_goal
)
# Evolve hypothesis based on feedback
evolved = evolution_agent.evolve_hypothesis(
hypothesis.get("hypothesis", hypothesis.get("statement")),
feedback,
self.research_goal
)
# Check proximity to research goal
proximity_result = proximity_agent.evaluate_proximity(
evolved.get("hypothesis", evolved.get("statement")),
self.research_goal
)
# Update hypothesis with refined version if it's relevant
if proximity_result["proximity_score"] >= 0.7: # Threshold for relevance
refined_hypotheses[j] = evolved
refined_hypotheses[j]["proximity"] = proximity_result
refined_hypotheses[j]["feedback"] = feedback
refined_hypotheses[j]["iteration"] = i + 1
main_text = evolved.get("hypothesis", evolved.get("statement", ""))
self.logger.info(f"Refined hypothesis {j+1}: {main_text[:100]}...")
else:
self.logger.warning(f"Refined hypothesis {j+1} was too far from research goal (score: {proximity_result['proximity_score']})")
# Update ranked hypotheses with refined versions
for i, refined in enumerate(refined_hypotheses):
self.ranked_hypotheses[i] = refined
return refined_hypotheses
def generate_research_report(self) -> Dict[str, Any]:
"""Generate a comprehensive research report with the findings.
Returns:
A dictionary containing the complete research report
Raises:
ValueError: If no hypotheses have been processed
"""
if not self.ranked_hypotheses:
raise ValueError("No hypotheses have been processed for report generation")
self.logger.info("Generating comprehensive research report")
# Get meta-review agent
meta_review_agent = self.agent_factory.get_agent("metareview")
# Generate report
report = meta_review_agent.generate_report(
self.research_goal,
self.ranked_hypotheses,
max_hypotheses=5 # Include top 5 hypotheses in report
)
self.final_report = report
self.logger.info(f"Generated research report with {report.get('hypothesis_count', 'unknown')} hypotheses")
return report
def save_results(self, output_dir: Optional[str] = None) -> str:
"""Save the research results to disk.
Args:
output_dir: Optional directory to save results. If None, saves to 'results'
directory in the project root.
Returns:
Path to the saved results file
Raises:
ValueError: If no final report is available
"""
if not self.final_report:
raise ValueError("No final report available to save")
# Import dependencies here to avoid circular imports
import time
from datetime import datetime
# Determine output directory
if output_dir is None:
# Get the project root directory (assuming this file is in src/)
project_root = Path(__file__).parent.parent # Go up one level from this file
output_dir = project_root / "results"
else:
output_dir = Path(output_dir)
# Create output directory if it doesn't exist
os.makedirs(output_dir, exist_ok=True)
# Create results dictionary
results = {
"research_goal": self.research_goal,
"configuration": self.config,
"hypotheses": {
"initial": self.hypotheses,
"ranked": self.ranked_hypotheses
},
"final_report": self.final_report,
"meta": {
"model": self.config["model"],
"timestamp": str(datetime.now().isoformat())
}
}
# Save results to file
results_file = output_dir / f"research_results_{time.strftime('%Y%m%d_%H%M%S')}.json"
with open(results_file, 'w') as f:
json.dump(results, f, indent=2)
self.logger.info(f"Saved research results to {results_file}")
return str(results_file)
def run_full_workflow(self, research_goal: str, iterations: int = 3, output_dir: Optional[str] = None) -> Dict[str, Any]:
"""Run the complete research workflow from goal to final report.
Args:
research_goal: The research goal to pursue
iterations: Number of refinement iterations
output_dir: Optional directory to save results
Returns:
The complete research results dictionary
Raises:
ValidationError: If research goal is invalid
"""
self.logger.info(f"Starting full research workflow for goal: {research_goal}")
try:
# Set research goal
self.set_research_goal(research_goal)
# Generate hypotheses
self.generate_hypotheses(count=7) # Generate 7 initial hypotheses
# Rank hypotheses
self.rank_hypotheses()
# Refine hypotheses
self.refine_hypotheses(iterations=iterations)
# Generate report
self.generate_research_report()
# Save results if output directory is specified
if output_dir:
self.save_results(output_dir)
self.logger.info("Research workflow completed successfully")
# Return final results
return {
"research_goal": self.research_goal,
"hypotheses": self.ranked_hypotheses,
"report": self.final_report
}
except Exception as e:
self.logger.error(f"Error in research workflow: {str(e)}")
raise
def main():
"""Main entry point for the AI Co-Scientist application."""
# For imports that may cause circular dependencies, import here
import time
from datetime import datetime
# Parse command line arguments
parser = argparse.ArgumentParser(description="AI Co-Scientist: Multi-agent system for scientific hypothesis generation")
parser.add_argument("--goal", "-g", type=str, help="Research goal to pursue")
parser.add_argument("--model", "-m", type=str, default=AGENT_DEFAULT_MODEL, help="LLM model to use")
parser.add_argument("--temp", "-t", type=float, default=AGENT_DEFAULT_TEMPERATURE, help="Temperature for LLM generation")
parser.add_argument("--iterations", "-i", type=int, default=3, help="Number of refinement iterations")
parser.add_argument("--output", "-o", type=str, help="Output directory for results")
parser.add_argument("--verbose", "-v", action="store_true", help="Enable verbose logging")
args = parser.parse_args()
# Configure logging level
log_level = "DEBUG" if args.verbose else LOG_LEVEL
logger = setup_logging(log_level=log_level, log_to_file=LOG_TO_FILE)
# Initialize configuration
config = {
"model": args.model,
"temperature": args.temp,
"max_iterations": MAX_ITERATIONS,
"max_tokens": MAX_TOKENS
}
# Initialize AI Co-Scientist
acs = AICoScientist(config=config)
# Run workflow if goal is provided
if args.goal:
try:
results = acs.run_full_workflow(
research_goal=args.goal,
iterations=args.iterations,
output_dir=args.output
)
# Print summary results
print("\n" + "=" * 80)
print(f"RESEARCH GOAL: {args.goal}")
print("=" * 80)
print("\nTOP HYPOTHESES:")
for i, h in enumerate(results["hypotheses"][:3]):
print(f"\n{i+1}. {h.get('hypothesis', h.get('statement', ''))}")
print(f" Score: {h.get('score', 'N/A')}")
print("\n" + "=" * 80)
print("EXECUTIVE SUMMARY:")
print(results["report"]["executive_summary"])
print("=" * 80 + "\n")
# Print path to full results if saved
if args.output:
print(f"Complete results saved to: {args.output}")
except Exception as e:
logger.error(f"Error in workflow execution: {str(e)}")
print(f"\nError: {str(e)}")
return 1
else:
parser.print_help()
return 0
if __name__ == "__main__":
main()
|