Datasets:

ArXiv:
File size: 6,821 Bytes
9f3bc09
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import json
import gradio as gr
from typing import List, Dict, Any
import pandas as pd

class AnnotationInterface:
    """Web interface for annotating plan agent trajectories"""
    
    def __init__(self, data_file: str):
        self.data = self.load_data(data_file)
        self.current_idx = 0
        self.annotations = []
    
    def load_data(self, file_path: str) -> List[Dict]:
        """Load trajectories for annotation"""
        with open(file_path, 'r') as f:
            data = json.load(f)
        return data.get("trajectories", [])
    
    def get_current_example(self) -> Dict:
        """Get current example for annotation"""
        if 0 <= self.current_idx < len(self.data):
            return self.data[self.current_idx]
        return {}
    
    def format_trajectory(self, trajectory: List[Dict]) -> str:
        """Format trajectory for display"""
        formatted = []
        for step in trajectory:
            if step["decision_type"] == "explore":
                formatted.append(f"Step {step['step_number']}:")
                formatted.append(f"  Sub-aspect: {step['sub_aspect']}")
                formatted.append(f"  Tool: {step['tool']}")
                formatted.append(f"  Thought: {step['thought']}")
            else:
                formatted.append(f"Final Summary:")
                formatted.append(f"  {step.get('summary', '')}")
        return "\n".join(formatted)
    
    def annotate_current(
        self,
        quality_score: int,
        strategy_appropriate: bool,
        exploration_complete: bool,
        optimal_stopping: bool,
        improvements: str,
        alternative_paths: str
    ) -> Dict:
        """Annotate current example"""
        annotation = {
            "example_idx": self.current_idx,
            "user_query": self.get_current_example().get("user_query", ""),
            "quality_score": quality_score,
            "strategy_appropriate": strategy_appropriate,
            "exploration_complete": exploration_complete,
            "optimal_stopping_point": optimal_stopping,
            "suggested_improvements": improvements,
            "alternative_exploration_paths": alternative_paths,
            "trajectory_length": len(self.get_current_example().get("trajectory", []))
        }
        
        self.annotations.append(annotation)
        return annotation
    
    def save_annotations(self, output_file: str = "annotations.json"):
        """Save all annotations"""
        with open(output_file, 'w') as f:
            json.dump({
                "total_annotations": len(self.annotations),
                "annotations": self.annotations
            }, f, indent=2)
        return f"Saved {len(self.annotations)} annotations"
    
    def create_interface(self):
        """Create Gradio interface"""
        with gr.Blocks() as interface:
            gr.Markdown("# Plan Agent Trajectory Annotation Tool")
            
            with gr.Row():
                with gr.Column(scale=2):
                    query_display = gr.Textbox(
                        label="User Query",
                        value=self.get_current_example().get("user_query", ""),
                        interactive=False
                    )
                    trajectory_display = gr.Textbox(
                        label="Exploration Trajectory",
                        value=self.format_trajectory(
                            self.get_current_example().get("trajectory", [])
                        ),
                        lines=20,
                        interactive=False
                    )
                
                with gr.Column(scale=1):
                    gr.Markdown("### Annotation")
                    quality_score = gr.Slider(
                        1, 5, value=3, step=1,
                        label="Overall Quality (1-5)"
                    )
                    strategy_appropriate = gr.Checkbox(
                        label="Strategy Appropriate for Query?"
                    )
                    exploration_complete = gr.Checkbox(
                        label="Exploration Sufficiently Complete?"
                    )
                    optimal_stopping = gr.Checkbox(
                        label="Stopped at Optimal Point?"
                    )
                    improvements = gr.Textbox(
                        label="Suggested Improvements",
                        lines=3
                    )
                    alternative_paths = gr.Textbox(
                        label="Alternative Exploration Paths",
                        lines=3
                    )
                    
                    with gr.Row():
                        prev_btn = gr.Button("Previous")
                        next_btn = gr.Button("Next")
                        save_btn = gr.Button("Save Annotations")
                    
                    progress = gr.Textbox(
                        label="Progress",
                        value=f"{self.current_idx + 1}/{len(self.data)}"
                    )
            
            # Button actions
            def go_next(q, s, e, o, i, a):
                self.annotate_current(q, s, e, o, i, a)
                self.current_idx = min(self.current_idx + 1, len(self.data) - 1)
                example = self.get_current_example()
                return (
                    example.get("user_query", ""),
                    self.format_trajectory(example.get("trajectory", [])),
                    f"{self.current_idx + 1}/{len(self.data)}"
                )
            
            def go_prev():
                self.current_idx = max(self.current_idx - 1, 0)
                example = self.get_current_example()
                return (
                    example.get("user_query", ""),
                    self.format_trajectory(example.get("trajectory", [])),
                    f"{self.current_idx + 1}/{len(self.data)}"
                )
            
            next_btn.click(
                go_next,
                inputs=[quality_score, strategy_appropriate, exploration_complete,
                       optimal_stopping, improvements, alternative_paths],
                outputs=[query_display, trajectory_display, progress]
            )
            
            prev_btn.click(
                go_prev,
                outputs=[query_display, trajectory_display, progress]
            )
            
            save_btn.click(
                lambda: self.save_annotations(),
                outputs=progress
            )
        
        return interface


# Usage
if __name__ == "__main__":
    # Create annotation interface
    annotator = AnnotationInterface("collected_trajectories.json")
    interface = annotator.create_interface()
    
    # Launch web interface
    interface.launch(share=True)