Hugging Face's logo

czty
/
Beyond_Prompt-based_Retrieval

Model card Files
xet
 Community
Beyond_Prompt-based_Retrieval
File size: 4,275 Bytes
a071401
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
# =========================================================
# Academic-style Ablation Figure
# =========================================================

import matplotlib.pyplot as plt
import numpy as np

configs = [
    "Biomni\n(Baseline)",
    "+MCP\nInfra.",
    "+MCP Infra.\n +Graph-Sca\n Planning",
]

# 仅更新此处的数字，完全对齐表格中的最新数据
metrics = {
    "BioAgentBench": [39.67, 42.68, 46.84],
    "LAB-Bench DbQA": [57.46, 66.35, 67.29],
    "LAB-Bench SeqQA": [84.76, 89.84, 90.48],
}

# =========================================================
# Muted academic palette
# =========================================================
fill_colors = [
    "#E5E7EB",   # soft gray
    "#D6E4F0",   # muted blue
    "#E8D6DC",   # muted wine
]

edge_colors = [
    "#6B7280",
    "#4B6A88",
    "#8A3D5D",
]

# =========================================================
# Style
# =========================================================
plt.rcParams.update(
    {
        "font.family": "DejaVu Sans",
        "font.size": 10,
        "axes.titlesize": 12,
        "xtick.labelsize": 9,
        "ytick.labelsize": 9.5,
        "pdf.fonttype": 42,
        "ps.fonttype": 42,
    }
)

fig, axes = plt.subplots(
    1,
    3,
    figsize=(8.4, 2.85),
    sharey=True,
)

y = np.arange(len(configs))

for idx, (ax, (title, values)) in enumerate(zip(axes, metrics.items())):

    values = np.array(values)
    baseline = values[0]

    span = values.max() - values.min()

    x_min = values.min() - span * 0.28 - 0.6
    x_max = values.max() + span * 0.52 + 1.2

    ax.set_xlim(x_min, x_max)

    # subtle baseline reference
    ax.axvline(
        baseline,
        color="#C7CCD4",
        linestyle="--",
        linewidth=0.9,
        zorder=1,
    )

    for i, val in enumerate(values):

        # main bar
        ax.barh(
            y[i],
            val - x_min,
            left=x_min,
            height=0.42,
            color=fill_colors[i],
            edgecolor=edge_colors[i],
            linewidth=1.15,
            zorder=3,
        )

        # subtle inner highlight
        ax.barh(
            y[i],
            (val - x_min) * 0.52,
            left=x_min,
            height=0.42,
            color="white",
            alpha=0.12,
            linewidth=0,
            zorder=4,
        )

        # value inside bar
        ax.text(
            val - span * 0.06,
            y[i],
            f"{val:.1f}",
            ha="right",
            va="center",
            fontsize=10,
            color=edge_colors[i],
            fontweight="bold" if i == 2 else "normal",
        )

        # gain
        if i > 0:
            gain = val - baseline

            ax.text(
                val + span * 0.14 + 0.2,
                y[i],
                f"+{gain:.1f}",
                ha="left",
                va="center",
                fontsize=8.8,
                color="#16924A",
            )

    # panel title
    ax.set_title(title, pad=7, fontweight="bold")

    ax.set_xticks([])

    # minimal style
    ax.spines["top"].set_visible(False)
    ax.spines["right"].set_visible(False)
    ax.spines["bottom"].set_visible(False)

    ax.spines["left"].set_color("#D1D5DB")
    ax.spines["left"].set_linewidth(0.8)

    if idx == 0:
        ax.set_yticks(y)
        ax.set_yticklabels(configs)
    else:
        ax.tick_params(axis="y", left=False, labelleft=False)

# top-to-bottom
axes[0].invert_yaxis()

# =========================================================
# Figure title
# =========================================================
fig.suptitle(
    "Ablation of MCP Infrastructure and Graph Planning",
    fontsize=13.5,
    fontweight="bold",
    y=1.02,
)

# =========================================================
# Layout
# =========================================================
fig.subplots_adjust(
    left=0.16,
    right=0.985,
    top=0.72,
    bottom=0.15,
    wspace=0.08,
)

# =========================================================
# Save
# =========================================================
fig.savefig("ablation_academic.pdf", bbox_inches="tight")
fig.savefig("ablation_academic.png", dpi=500, bbox_inches="tight")
fig.savefig("ablation_academic.svg", bbox_inches="tight")