continuation_benchmark/heldout_eng_r3/summary.json

{
  "version": "heldout_eng_r3",
  "generated_utc": "2026-05-02T22:27:51.487211Z",
  "datasets": {
    "F1_seva_silva_rocha": "JX560290:JX560360[ACCN] AND plasmid[All Fields]",
    "F2_broad_host_range": "broad host range vector[Title] AND plasmid[Title] NOT Addgene[All Fields] AND (1500[SLEN] : 30000[SLEN])",
    "F3_pseudomonas_2024": "Pseudomonas[Organism] AND plasmid[Title] AND (\"2024/04/01\"[PDAT] : \"2026/12/31\"[PDAT]) AND (1500[SLEN] : 30000[SLEN])",
    "F4_seva_niche_origins": "(pBBR1[All Fields] OR RK2[All Fields] OR RSF1010[All Fields] OR pSC101[All Fields]) AND plasmid[Title] AND vector[Title] NOT Addgene[All Fields] AND (1500[SLEN] : 30000[SLEN])",
    "F5_yarrowia_rhodococcus": "(Yarrowia[Organism] OR Rhodococcus[Organism]) AND plasmid[Title] AND (1500[SLEN] : 30000[SLEN])",
    "F6_bacillus_2024plus": "Bacillus[Organism] AND plasmid[Title] AND (\"2024/04/01\"[PDAT] : \"2026/12/31\"[PDAT]) AND (1500[SLEN] : 30000[SLEN])"
  },
  "completion": {
    "total_windows": 25223,
    "by_dataset": {
      "F3_pseudomonas_2024": {
        "Base": {
          "mean": -7.9114,
          "std": 0.9525,
          "n": 11066
        },
        "SFT": {
          "mean": -7.7731,
          "std": 0.9766,
          "n": 11066
        },
        "RL": {
          "mean": -8.0938,
          "std": 0.7639,
          "n": 11066
        }
      },
      "F4_seva_niche_origins": {
        "Base": {
          "mean": -5.1206,
          "std": 4.1919,
          "n": 75
        },
        "SFT": {
          "mean": -4.4735,
          "std": 4.2419,
          "n": 75
        },
        "RL": {
          "mean": -6.818,
          "std": 3.2432,
          "n": 75
        }
      },
      "F5_yarrowia_rhodococcus": {
        "Base": {
          "mean": -7.7796,
          "std": 0.7823,
          "n": 4571
        },
        "SFT": {
          "mean": -7.8611,
          "std": 0.8056,
          "n": 4571
        },
        "RL": {
          "mean": -7.9882,
          "std": 0.6742,
          "n": 4571
        }
      },
      "F6_bacillus_2024plus": {
        "Base": {
          "mean": -7.8873,
          "std": 0.9162,
          "n": 9466
        },
        "SFT": {
          "mean": -7.6428,
          "std": 0.7406,
          "n": 9466
        },
        "RL": {
          "mean": -8.0285,
          "std": 0.673,
          "n": 9466
        }
      }
    }
  },
  "surprisal": {
    "total_windows": 8000,
    "by_dataset": {
      "F5_yarrowia_rhodococcus": {
        "Base": {
          "mean": -12.5404,
          "std": 1.3053,
          "n": 1425
        },
        "SFT": {
          "mean": -8.694,
          "std": 0.8361,
          "n": 1425
        },
        "RL": {
          "mean": -9.9298,
          "std": 0.8831,
          "n": 1425
        }
      },
      "F3_pseudomonas_2024": {
        "Base": {
          "mean": -12.5858,
          "std": 1.3024,
          "n": 3612
        },
        "SFT": {
          "mean": -8.6216,
          "std": 0.8803,
          "n": 3612
        },
        "RL": {
          "mean": -10.0258,
          "std": 0.9308,
          "n": 3612
        }
      },
      "F6_bacillus_2024plus": {
        "Base": {
          "mean": -12.4792,
          "std": 1.3861,
          "n": 2937
        },
        "SFT": {
          "mean": -8.2548,
          "std": 0.7883,
          "n": 2937
        },
        "RL": {
          "mean": -9.8981,
          "std": 0.8754,
          "n": 2937
        }
      },
      "F4_seva_niche_origins": {
        "Base": {
          "mean": -13.0092,
          "std": 2.688,
          "n": 26
        },
        "SFT": {
          "mean": -7.3158,
          "std": 2.5211,
          "n": 26
        },
        "RL": {
          "mean": -10.7329,
          "std": 1.9973,
          "n": 26
        }
      }
    }
  }
}