Update CogniRank with SmolScript and refreshed model rankings

Add the five-benchmark CogniRank update with SmolScript at 20%, immutable source/model pins, and compatible completed measurements for all 15 models. Preserve the CogniRank public name.

README.md

@@ -12,4 +12,10 @@ short_description: leaderboard that measures cognitive capacity on small models
 license: cc-by-nc-nd-4.0
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# CogniRank
+
+The November update measures five capabilities in small language models. SmolCUA contributes 24/95 (25.26%), MARC 20/95 (21.05%), Procedural Pile 20/95 (21.05%), TTM 12/95 (12.63%), and SmolScript 19/95 (20%). All metrics use a 0–100 scale. An overall index requires all five completed components.
+
+[Independent evaluator, exact dataset/model revisions, prompts, sandbox setup, and result provenance](https://huggingface.co/datasets/dotlabs/cognirank). SmolScript uses 350 fixed examples across nine languages and macro-averaged execution Pass@1. Published measurements use a requested context of 4,096; context extensions are experiments, and unsupported fixed contexts receive no index.
+
+`data.json` is authoritative; the HTML includes a matching offline snapshot. Previous four-component indices have been replaced using completed compatible SmolScript runs, without reinterpreting older SmolScript protocols.

data.json

@@ -1,10 +1,11 @@
 {
-  "schema_version": 15,
+  "schema_version": 16,
   "evaluation_weights": {
-    "ttm": 0.15789473684210525,
-    "marc": 0.2631578947368421,
-    "procedural_pile": 0.2631578947368421,
-    "smolcua": 0.3157894736842105
+    "smolcua": 0.25263157894736843,
+    "marc": 0.21052631578947367,
+    "procedural_pile": 0.21052631578947367,
+    "ttm": 0.12631578947368421,
+    "smolscript": 0.2
   },
   "eligibility": {
     "min_parameters_millions": 25,
@@ -35,7 +36,21 @@
             "ttm": 42.55,
             "marc": 11.5625,
             "procedural_pile": 0.028325667131668543,
-            "smolcua": 0.0
+            "smolcua": 0.0,
+            "smolscript": 0.0
+          },
+          "cognirank_index": 7.814910666764562,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "d9d2d7c6abb47f86fed6c87cf95e189f4686a7705fc8f3f3cb16b7b6dcc0958a",
+              "smolscript": "816dda142e9c336adace796e59bc04e139c0472ee729c12d50edf67913ac8032"
+            }
           }
         }
       ]
@@ -62,7 +77,21 @@
             "ttm": 46.7,
             "marc": 5.3125,
             "procedural_pile": 0.004785877737822812,
-            "smolcua": 0.0
+            "smolcua": 0.0,
+            "smolscript": 0.5374255374255374
+          },
+          "cognirank_index": 7.125861081745702,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "b863c457cb2b5d6f32650f8a55b0cf90df150391aca002c923eef1473ceb054d",
+              "smolscript": "b4e8f5fa96434823307edee7e14b66fe8b9fc9be048ebcf7c00800ae4474270a"
+            }
           }
         }
       ]
@@ -89,7 +118,21 @@
             "ttm": 43.9,
             "marc": 6.25,
             "procedural_pile": 0.0034782608695652162,
-            "smolcua": 0.0
+            "smolcua": 0.0,
+            "smolscript": 0.2849002849002849
+          },
+          "cognirank_index": 6.918764954005228,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "89c8ea23447a276d4394e157f7b4bd58778545eb1d6b84c8759b161f4379cdb5",
+              "smolscript": "acfee735d0491c330985a3a1b38a75281815f99be4418a9e0acf9e97986aecac"
+            }
           }
         }
       ]
@@ -116,7 +159,21 @@
             "ttm": 41.65,
             "marc": 1.875,
             "procedural_pile": 0.0426930028759097,
-            "smolcua": 0.0
+            "smolcua": 0.0,
+            "smolscript": 0.0
+          },
+          "cognirank_index": 5.664777474289665,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "4a79299d13bba736d8f82251d7ab77044af3513b3a5ccdbdf601a89bbdde4926",
+              "smolscript": "eb0e2a6253c4158e035e7da9598252f2ac43ed61355e012472ced6d5c8ae2470"
+            }
           }
         }
       ]
@@ -143,7 +200,21 @@
             "ttm": 43.3,
             "marc": 7.1875,
             "procedural_pile": 7.969346397750256,
-            "smolcua": 0.0
+            "smolcua": 0.0,
+            "smolscript": 0.8223258223258223
+          },
+          "cognirank_index": 8.824853879781008,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "d6a2bdc3501fb7e56855aff8e79a117b33e5b57aec39b814ed97f65f1e6ef85b",
+              "smolscript": "570af0a153d49e39125d61cee1cea5368e2425d6d6aad87306c3a7f7a0c6f1fb"
+            }
           }
         }
       ]
@@ -170,7 +241,21 @@
             "ttm": 48.2,
             "marc": 0.0,
             "procedural_pile": 2.780199447600087,
-            "smolcua": 8.1778
+            "smolcua": 8.1778,
+            "smolscript": 0.0
+          },
+          "cognirank_index": 8.739696725810544,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "85f3542c267a43e3ee0eb7af66e2f4ce046b4d25e6f15ee2d2acbd9820f5868b",
+              "smolscript": "8e6040d0e2c4e9b6e8c4af8fe9fda4229250cbfb4b0ef9c62b38edd916e3c749"
+            }
           }
         },
         {
@@ -189,7 +274,21 @@
             "ttm": 46.85,
             "marc": 0.0,
             "procedural_pile": 11.993131773709633,
-            "smolcua": 14.0053
+            "smolcua": 14.0053,
+            "smolscript": 0.0
+          },
+          "cognirank_index": 11.98094563657045,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "23cf0691a4f482164e83c77ca277ba4102c8ea85bd3db222bf3f2f6e9ec7fdce",
+              "smolscript": "3a9b27124c8417e6217d3d881bd3aa683a940c9b0b68092b2e03e52a471681fa"
+            }
           }
         }
       ]
@@ -216,7 +315,21 @@
             "ttm": 46.1,
             "marc": 2.8125,
             "procedural_pile": 0.0029742541128357264,
-            "smolcua": 0.0
+            "smolcua": 0.0,
+            "smolscript": 0.0
+          },
+          "cognirank_index": 6.415889316655334,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "6e94f6f570384bb6e5af5e52079a1016e66041d5d162e492b255451200ed8590",
+              "smolscript": "b8f878ea66e57eed40b3bd22760aeff04d6b2ed26aaebd252043da516f2f21be"
+            }
           }
         }
       ]
@@ -236,14 +349,28 @@
           "evaluation_status": "evaluated",
           "date_submitted": "2026-11-04",
           "date_added": "2026-11-04",
-          "author_note": "IBM’s compact, creator-post-trained Granite model for practical language, code, and tool-use tasks.",
+          "author_note": "IBM\u2019s compact, creator-post-trained Granite model for practical language, code, and tool-use tasks.",
           "comment_source": "dotlabs",
           "checkpoint_revision": "bd8a1497065c0d6ba1ef19af6b0d2b14bacf71c2",
           "scores": {
             "ttm": 52.7,
             "marc": 2.5,
             "procedural_pile": 13.301300628673525,
-            "smolcua": 13.8532
+            "smolcua": 13.8532,
+            "smolscript": 12.960095460095461
+          },
+          "cognirank_index": 16.075206592792465,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "2a230917ac62150d383e5641c66324138679c14ed29d323747951f7d1cfa6874",
+              "smolscript": "56fd716f2cf690e2a7fd723cbabae9dd783a2c85b8bbd2b482c67cec8375d8dd"
+            }
           }
         },
         {
@@ -255,14 +382,28 @@
           "evaluation_status": "evaluated",
           "date_submitted": "2026-11-04",
           "date_added": "2026-11-04",
-          "author_note": "IBM’s hybrid attention–Mamba2 Granite model, post-trained for efficient instruction and tool-oriented tasks.",
+          "author_note": "IBM\u2019s hybrid attention\u2013Mamba2 Granite model, post-trained for efficient instruction and tool-oriented tasks.",
           "comment_source": "dotlabs",
           "checkpoint_revision": "3b17b717b8f2f5d305b0a92c1491e239aeda19c8",
           "scores": {
             "ttm": 51.4,
             "marc": 3.125,
             "procedural_pile": 15.574583666404257,
-            "smolcua": 15.9499
+            "smolcua": 15.9499,
+            "smolscript": 13.165815665815666
+          },
+          "cognirank_index": 17.09199758924824,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "055a122c6216e54ee46d138f3e4e8404a5b8403fed73239bd14e8467b6f20698",
+              "smolscript": "439c60e134c754cfa18b1f4a9b04deab4021a2d29a24baaeebd20fb476a519d8"
+            }
           }
         }
       ]
@@ -289,7 +430,21 @@
             "ttm": 43.55,
             "marc": 3.4375,
             "procedural_pile": 0.012771832822402176,
-            "smolcua": 0
+            "smolcua": 0,
+            "smolscript": 0.0
+          },
+          "cognirank_index": 6.227425649015242,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "3966ed32879da5f22c99cd5b39c4536917af385110688bb5ee4ca1651e4d3d91",
+              "smolscript": "a5fa7426a9801dde80cffc21bed32cd5caa27e23e2ea66aefb30bdd6b10e065f"
+            }
           }
         },
         {
@@ -308,7 +463,21 @@
             "ttm": 45.6,
             "marc": 6.875,
             "procedural_pile": 0.027063257654258682,
-            "smolcua": 0
+            "smolcua": 0,
+            "smolscript": 0.854885854885855
+          },
+          "cognirank_index": 7.384043119957015,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "b303d6e3b8599a2daf7c5cce2f4302b47def18ffd0ac340292c083aea477c28a",
+              "smolscript": "098282cdd78b1230f5d882d213c51eeb1cfd80b22d431a3bedf06378856ac2a9"
+            }
           }
         }
       ]
@@ -335,7 +504,21 @@
             "ttm": 40.2,
             "marc": 4.0625,
             "procedural_pile": 0.004968944099378885,
-            "smolcua": 0
+            "smolcua": 0,
+            "smolscript": 0.25252525252525254
+          },
+          "cognirank_index": 5.9847090387365,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "85fc0e5e1a94c38ee36c0ced342248c8bfe4be30cea1a0b9b4d6af1855361247",
+              "smolscript": "ef0952687e5fe1268f811433da950c461dc57480fc23f9bbb71debaaca830829"
+            }
           }
         }
       ]
@@ -362,7 +545,21 @@
             "ttm": 44.6,
             "marc": 10.625,
             "procedural_pile": 0.019425921663509163,
-            "smolcua": 0
+            "smolcua": 0,
+            "smolscript": 0.2849002849002849
+          },
+          "cognirank_index": 7.931596040488165,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "006a655293b27ae8efe5576db5cb8a03aed1f45e83d4031fe8be42c4114e600b",
+              "smolscript": "97eada25f949017706f31f8d48c23a98fd3a1b2172fbe787ebe8dd3ce9a4c145"
+            }
           }
         },
         {
@@ -381,10 +578,28 @@
             "ttm": 46.55,
             "marc": 7.1875,
             "procedural_pile": 0.007577656444398657,
-            "smolcua": 0
+            "smolcua": 0,
+            "smolscript": 0.0
+          },
+          "cognirank_index": 7.394753190830399,
+          "evaluation": {
+            "protocol": "cognirank-2026-11-update",
+            "requested_context": 4096,
+            "component_protocols": {
+              "reasoning": "cognirank-v9",
+              "smolscript": "smolscript-0.5.0"
+            },
+            "source_result_sha256": {
+              "reasoning": "dfd91f864c36ca118a014a240e4a0fe80d79b318930d84399f9c1459a5bd5caf",
+              "smolscript": "23f98009db8156346cd763c518bf4ed9d821575bc956b88471ea219cea04d51b"
+            }
           }
         }
       ]
     }
-  ]
+  ],
+  "evaluation_protocol": "cognirank-2026-11-update",
+  "updated_at": "2026-11-05",
+  "reproduction_url": "https://huggingface.co/datasets/dotlabs/cognirank",
+  "evaluation_context": 4096
 }

index.html

@@ -96,7 +96,7 @@ button,input{font:inherit}button{cursor:pointer}button:focus-visible,input:focus
 .pane{min-width:0;min-height:0;border:1px solid var(--line);border-radius:16px;display:flex;flex-direction:column;background:#000;overflow:hidden}
 .list-scroller{min-height:0;flex:1;overflow:auto;overscroll-behavior:contain;scrollbar-width:thin;scrollbar-color:#454545 #000}
 table{border-collapse:collapse;width:100%;table-layout:fixed}
-col.rank{width:45px}col.model{width:270px}col.total{width:78px}col.metric{width:65px}
+col.rank{width:45px}col.model{width:245px}col.total{width:78px}col.metric{width:62px}
 th{position:sticky;top:0;z-index:1;background:#000;color:#777;font-size:10px;font-weight:500;text-align:left;height:43px;white-space:nowrap;border-bottom:1px solid var(--line)}
 th:first-child,td:first-child{padding-left:17px}th:last-child,td:last-child{padding-right:12px}th.number,td.number{text-align:center}
 th[data-sort]{cursor:pointer;user-select:none}th[data-sort]:hover,th[data-sort].active-sort{color:#f2f2f2}
@@ -163,7 +163,7 @@ tbody tr.virtual-space td{height:0;min-height:0;padding:0;border:0;line-height:0
 @media(hover:none){.chart-inspect{display:none}}
 
 .notice{position:absolute;z-index:5;left:50%;bottom:12px;transform:translate(-50%,10px);background:#fff;color:#000;border-radius:8px;padding:9px 12px;font-size:12px;opacity:0;pointer-events:none;transition:opacity .2s,transform .2s;white-space:nowrap}.notice.show{opacity:1;transform:translate(-50%,0)}
-@media(max-width:1150px){col.metric{width:52px}col.model{width:258px}.board{grid-template-columns:minmax(0,1.55fr) minmax(275px,.9fr)}.app.compare .board{grid-template-columns:minmax(260px,.65fr) minmax(0,1.35fr)}}
+@media(max-width:1150px){col.metric{width:52px}col.model{width:210px}.board{grid-template-columns:minmax(0,1.55fr) minmax(275px,.9fr)}.app.compare .board{grid-template-columns:minmax(260px,.65fr) minmax(0,1.35fr)}}
 @media(max-width:980px) and (min-width:651px){.size-filter{gap:3px}.size-filter svg{width:12px}.model-heading-inner{gap:4px}}
 @media(max-width:850px){.app.compare .size-filter{position:static;flex:1 1 auto;width:auto;max-width:100%}.board{grid-template-columns:minmax(0,1.3fr) minmax(270px,1fr)}col.metric,th.metric,td.metric{display:none}col.model{width:auto}.app.compare .board{grid-template-columns:minmax(230px,.65fr) minmax(0,1.35fr)}}
 @media(max-width:650px){.app{padding:11px 12px 12px}.top{height:78px;gap:6px}.brand{font-size:12px;margin-top:16px}.name{font-size:clamp(23px,7vw,31px)}.topright{font-size:9px;padding-top:19px}.controls{min-height:84px;grid-template-columns:minmax(0,1fr);grid-template-rows:38px 37px;gap:3px}.tabs{grid-row:1;grid-column:1}.size-filter{height:29px;gap:5px;max-width:100%;flex:1 1 180px}.size-filter svg{width:12px}.search-tools{grid-row:2;grid-column:1;justify-self:stretch;width:100%;max-width:none}.search{width:100%;max-width:none;min-width:0}.contribute{grid-row:1;grid-column:1;font-size:10px;padding:7px 9px}.board,.app.compare .board{display:block;position:relative}.pane{position:absolute;inset:0}.app:not(.compare) .chart-pane{display:none}.app.compare .list-pane{display:flex;bottom:auto;height:39%}.app.compare .chart-pane{top:calc(39% + 8px);bottom:0}td{height:54px}.chart{padding-top:5px}.legend{min-height:65px}}
@@ -308,12 +308,12 @@ body,button,input{font-family:var(--font-interface)}
     <main class="board">
       <section class="pane list-pane" aria-label="Model rankings">
         <div class="list-scroller" id="list-scroller" role="region" aria-label="Scrollable model rankings" tabindex="0">
-          <table aria-label="CogniRank model rankings"><colgroup><col class="rank"><col class="model"><col class="total"><col class="metric"><col class="metric"><col class="metric"><col class="metric"></colgroup>
+          <table aria-label="CogniRank model rankings"><colgroup><col class="rank"><col class="model"><col class="total"><col class="metric"><col class="metric"><col class="metric"><col class="metric"><col class="metric"></colgroup>
           <thead><tr><th>#</th><th class="model-heading"><div class="model-heading-inner"><span>Model</span><div class="size-filter" title="Filter by parameter count">
         <span class="size-label" id="size-label" aria-hidden="true">&lt;350M</span>
         <svg viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="1.7" aria-hidden="true"><path d="M4 7h16M4 17h16"/><circle cx="9" cy="7" r="2.2" fill="#000"/><circle cx="16" cy="17" r="2.2" fill="#000"/></svg>
         <div class="size-track" id="size-track"><label for="size-range" class="visually-hidden">Maximum model size</label><input type="range" id="size-range" min="0" max="6" step="1" value="6" aria-valuetext="Under 350 million parameters"></div>
-      </div></div></th><th class="number active-sort" data-sort="score" aria-sort="descending">Score</th><th class="number metric" data-sort="ttm" title="TTM, 15.79%">TTM</th><th class="number metric" data-sort="marc" title="MARC, 26.32%">MARC</th><th class="number metric" data-sort="proc" title="Procedural Pile, 26.32%">ProcPile</th><th class="number metric" data-sort="smol" title="SmolCUA, 31.58%">Smol</th></tr></thead>
+      </div></div></th><th class="number active-sort" data-sort="score" aria-sort="descending">Score</th><th class="number metric" data-sort="ttm" title="TTM, 12.63%">TTM</th><th class="number metric" data-sort="marc" title="MARC, 21.05%">MARC</th><th class="number metric" data-sort="proc" title="Procedural Pile, 21.05%">ProcPile</th><th class="number metric" data-sort="smol" title="SmolCUA, 25.26%">Smol</th><th class="number metric" data-sort="code" title="SmolScript, 20%">Script</th></tr></thead>
           <tbody id="rows"></tbody></table><div class="empty" id="empty">No matches.</div>
         </div>
       </section>
@@ -334,7 +334,7 @@ body,button,input{font-family:var(--font-interface)}
     </section>
     <section class="update-slide" aria-hidden="true" inert>
       <img class="update-art" src="data:image/svg+xml;base64,<svg xmlns="http://www.w3.org/2000/svg" width="960" height="390" viewBox="0 0 960 390">
<rect width="960" height="390" fill="#000"/>
<circle cx="70.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="90.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="109.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="129.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="148.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="168.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="187.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="70.5" cy="154.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="187.5" cy="154.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="70.5" cy="173.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="187.5" cy="173.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="70.5" cy="193.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="187.5" cy="193.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="70.5" cy="212.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="187.5" cy="212.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="70.5" cy="232.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="187.5" cy="232.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="70.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="90.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="109.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="129.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="148.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="168.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="187.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="90.0" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="109.5" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="129.0" cy="154.0" r="5" fill="#fff" opacity=".85"/>
<circle cx="148.5" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="168.0" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="90.0" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="109.5" cy="173.5" r="5" fill="#fff" opacity=".85"/>
<circle cx="129.0" cy="173.5" r="5" fill="#fff" opacity=".85"/>
<circle cx="148.5" cy="173.5" r="5" fill="#fff" opacity=".85"/>
<circle cx="168.0" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="90.0" cy="193.0" r="5" fill="#fff" opacity=".85"/>
<circle cx="109.5" cy="193.0" r="5" fill="#fff" opacity=".85"/>
<circle cx="129.0" cy="193.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="148.5" cy="193.0" r="5" fill="#fff" opacity=".85"/>
<circle cx="168.0" cy="193.0" r="5" fill="#fff" opacity=".85"/>
<circle cx="90.0" cy="212.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="109.5" cy="212.5" r="5" fill="#fff" opacity=".85"/>
<circle cx="129.0" cy="212.5" r="5" fill="#fff" opacity=".85"/>
<circle cx="148.5" cy="212.5" r="5" fill="#fff" opacity=".85"/>
<circle cx="168.0" cy="212.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="90.0" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="109.5" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="129.0" cy="232.0" r="5" fill="#fff" opacity=".85"/>
<circle cx="148.5" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="168.0" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<path d="M86 295H172" stroke="#303030" stroke-width="1.5"/>
<circle cx="304.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="324.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="343.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="363.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="382.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="402.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="421.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="304.5" cy="154.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="421.5" cy="154.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="304.5" cy="173.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="421.5" cy="173.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="304.5" cy="193.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="421.5" cy="193.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="304.5" cy="212.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="421.5" cy="212.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="304.5" cy="232.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="421.5" cy="232.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="304.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="324.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="343.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="363.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="382.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="402.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="421.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="324.0" cy="154.0" r="5" fill="#fff" opacity=".95"/>
<circle cx="343.5" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="363.0" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="382.5" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="402.0" cy="154.0" r="5" fill="#fff" opacity=".95"/>
<circle cx="324.0" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="343.5" cy="173.5" r="5" fill="#fff" opacity=".95"/>
<circle cx="363.0" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="382.5" cy="173.5" r="5" fill="#fff" opacity=".95"/>
<circle cx="402.0" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="324.0" cy="193.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="343.5" cy="193.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="363.0" cy="193.0" r="5" fill="#fff" opacity=".95"/>
<circle cx="382.5" cy="193.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="402.0" cy="193.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="324.0" cy="212.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="343.5" cy="212.5" r="5" fill="#fff" opacity=".95"/>
<circle cx="363.0" cy="212.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="382.5" cy="212.5" r="5" fill="#fff" opacity=".95"/>
<circle cx="402.0" cy="212.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="324.0" cy="232.0" r="5" fill="#fff" opacity=".95"/>
<circle cx="343.5" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="363.0" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="382.5" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="402.0" cy="232.0" r="5" fill="#fff" opacity=".95"/>
<path d="M320 295H406" stroke="#303030" stroke-width="1.5"/>
<circle cx="538.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="558.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="577.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="597.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="616.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="636.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="655.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="538.5" cy="154.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="655.5" cy="154.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="538.5" cy="173.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="655.5" cy="173.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="538.5" cy="193.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="655.5" cy="193.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="538.5" cy="212.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="655.5" cy="212.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="538.5" cy="232.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="655.5" cy="232.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="538.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="558.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="577.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="597.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="616.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="636.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="655.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="558.0" cy="154.0" r="5" fill="#fff" opacity=".90"/>
<circle cx="577.5" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="597.0" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="616.5" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="636.0" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="558.0" cy="173.5" r="5" fill="#fff" opacity=".90"/>
<circle cx="577.5" cy="173.5" r="5" fill="#fff" opacity=".90"/>
<circle cx="597.0" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="616.5" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="636.0" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="558.0" cy="193.0" r="5" fill="#fff" opacity=".90"/>
<circle cx="577.5" cy="193.0" r="5" fill="#fff" opacity=".90"/>
<circle cx="597.0" cy="193.0" r="5" fill="#fff" opacity=".90"/>
<circle cx="616.5" cy="193.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="636.0" cy="193.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="558.0" cy="212.5" r="5" fill="#fff" opacity=".90"/>
<circle cx="577.5" cy="212.5" r="5" fill="#fff" opacity=".90"/>
<circle cx="597.0" cy="212.5" r="5" fill="#fff" opacity=".90"/>
<circle cx="616.5" cy="212.5" r="5" fill="#fff" opacity=".90"/>
<circle cx="636.0" cy="212.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="558.0" cy="232.0" r="5" fill="#fff" opacity=".90"/>
<circle cx="577.5" cy="232.0" r="5" fill="#fff" opacity=".90"/>
<circle cx="597.0" cy="232.0" r="5" fill="#fff" opacity=".90"/>
<circle cx="616.5" cy="232.0" r="5" fill="#fff" opacity=".90"/>
<circle cx="636.0" cy="232.0" r="5" fill="#fff" opacity=".90"/>
<path d="M554 295H640" stroke="#303030" stroke-width="1.5"/>
<circle cx="772.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="792.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="811.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="831.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="850.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="870.0" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="889.5" cy="134.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="772.5" cy="154.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="889.5" cy="154.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="772.5" cy="173.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="889.5" cy="173.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="772.5" cy="193.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="889.5" cy="193.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="772.5" cy="212.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="889.5" cy="212.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="772.5" cy="232.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="889.5" cy="232.0" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="772.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="792.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="811.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="831.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="850.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="870.0" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="889.5" cy="251.5" r="1.45" fill="#fff" opacity=".16"/>
<circle cx="792.0" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="811.5" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="831.0" cy="154.0" r="5" fill="#fff" opacity="1"/>
<circle cx="850.5" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="870.0" cy="154.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="792.0" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="811.5" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="831.0" cy="173.5" r="5" fill="#fff" opacity="1"/>
<circle cx="850.5" cy="173.5" r="5" fill="#fff" opacity="1"/>
<circle cx="870.0" cy="173.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="792.0" cy="193.0" r="5" fill="#fff" opacity="1"/>
<circle cx="811.5" cy="193.0" r="5" fill="#fff" opacity="1"/>
<circle cx="831.0" cy="193.0" r="5" fill="#fff" opacity="1"/>
<circle cx="850.5" cy="193.0" r="5" fill="#fff" opacity="1"/>
<circle cx="870.0" cy="193.0" r="5" fill="#fff" opacity="1"/>
<circle cx="792.0" cy="212.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="811.5" cy="212.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="831.0" cy="212.5" r="5" fill="#fff" opacity="1"/>
<circle cx="850.5" cy="212.5" r="5" fill="#fff" opacity="1"/>
<circle cx="870.0" cy="212.5" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="792.0" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="811.5" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="831.0" cy="232.0" r="5" fill="#fff" opacity="1"/>
<circle cx="850.5" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<circle cx="870.0" cy="232.0" r="2.1" fill="#fff" opacity="0.21"/>
<path d="M788 295H874" stroke="#303030" stroke-width="1.5"/>
</svg>" alt="" draggable="false" width="960" height="390">
-      <div class="update-text"><time class="update-date" datetime="2026-11-04">Nov 4, 2026</time><h3>Four complementary capabilities.</h3><p>The leaderboard focuses on social, abstract, procedural, and agentic reasoning—not perplexity or language fluency.</p></div>
+      <div class="update-text"><time class="update-date" datetime="2026-11-05">Nov 5, 2026</time><h3>CogniRank update: code joins the mix.</h3><p>SmolScript now contributes 20% across nine programming languages. Rankings use the updated five-benchmark formula.</p></div>
     </section>
   </div>
   <nav class="update-navigation" aria-label="Updates carousel">
@@ -349,7 +349,7 @@ body,button,input{font-family:var(--font-interface)}
     <section class="story-slide is-current" aria-hidden="false">
       <span class="story-kicker">COGNIRANK</span>
       <h3>What small models can do.</h3>
-      <p class="story-copy">CogniRank measures problem-solving in compact models, including first-party post-trained variants: social reasoning, abstraction, procedural reasoning, and action decisions.</p>
+      <p class="story-copy">CogniRank measures problem-solving in compact models, including first-party post-trained variants: social reasoning, abstraction, procedural reasoning, action decisions, and basic programming.</p>
     </section>
     <section class="story-slide" aria-hidden="true" inert>
       <span class="story-kicker">DOTLABS</span>
@@ -375,8 +375,8 @@ body,button,input{font-family:var(--font-interface)}
     </section>
     <section class="story-slide" aria-hidden="true" inert>
       <span class="story-kicker">EVALUATION</span>
-      <h3>Four measurable skills.</h3>
-      <p class="story-copy"><a class="benchmark-link" href="https://huggingface.co/datasets/AxiomicLabs/Tiny_Theory_of_Mind" target="_blank" rel="noopener noreferrer" aria-label="TTM benchmark on Hugging Face (opens in new tab)">TTM</a>, <a class="benchmark-link" href="https://huggingface.co/spaces/FlameF0X/Micro-ARC" target="_blank" rel="noopener noreferrer" aria-label="MARC benchmark on Hugging Face (opens in new tab)">MARC</a>, <a class="benchmark-link" href="https://huggingface.co/datasets/reasoning-core/procedural-pile" target="_blank" rel="noopener noreferrer" aria-label="Procedural Pile dataset on Hugging Face (opens in new tab)">Procedural Pile</a>, and <a class="benchmark-link" href="https://huggingface.co/datasets/dotlabs/smolcua" target="_blank" rel="noopener noreferrer" aria-label="SmolCUA dataset on Hugging Face (opens in new tab)">SmolCUA</a>. Zero-shot by default; MARC keeps its built-in examples. Each task is scored separately and combined when all results are available.</p>
+      <h3>Five measurable skills.</h3>
+      <p class="story-copy"><a class="benchmark-link" href="https://huggingface.co/datasets/AxiomicLabs/Tiny_Theory_of_Mind" target="_blank" rel="noopener noreferrer" aria-label="TTM benchmark on Hugging Face (opens in new tab)">TTM</a>, <a class="benchmark-link" href="https://huggingface.co/spaces/FlameF0X/Micro-ARC" target="_blank" rel="noopener noreferrer" aria-label="MARC benchmark on Hugging Face (opens in new tab)">MARC</a>, <a class="benchmark-link" href="https://huggingface.co/datasets/reasoning-core/procedural-pile" target="_blank" rel="noopener noreferrer" aria-label="Procedural Pile dataset on Hugging Face (opens in new tab)">Procedural Pile</a>, and <a class="benchmark-link" href="https://huggingface.co/datasets/dotlabs/smolcua" target="_blank" rel="noopener noreferrer" aria-label="SmolCUA dataset on Hugging Face (opens in new tab)">SmolCUA</a>, and <a class="benchmark-link" href="https://huggingface.co/datasets/dotlabs/smolscript" target="_blank" rel="noopener noreferrer">SmolScript</a>. Weights: SmolCUA 25.26%, MARC 21.05%, Procedural Pile 21.05%, TTM 12.63%, SmolScript 20%. SmolScript uses 350 fixed examples and plain code-completion prompts. MARC retains its demonstrations. All five scores are required. <a class="benchmark-link" href="https://huggingface.co/datasets/dotlabs/cognirank" target="_blank" rel="noopener noreferrer">Reproduce the update</a>.</p>
     </section>
     <section class="story-slide" aria-hidden="true" inert>
       <span class="story-kicker">SUBMISSION</span>
@@ -421,10 +421,10 @@ body,button,input{font-family:var(--font-interface)}
 // Brand colors are read from data.json; deterministic dot patterns need no image assets.
 const previewAuthors={};
 const localPreviewRepos={};
-const axes=[['ttm','Social',.15/.95],['marc','Abstract',.25/.95],['proc','Procedural',.25/.95],['smol','Agentic',.30/.95]];
+const axes=[['ttm','Social',12/95],['marc','Abstract',20/95],['proc','Procedural',20/95],['smol','Agentic',24/95],['code','Programming',19/95]];
 let models=[];
 let map=new Map();
-const scoreJsonKeys={smol:'smolcua',marc:'marc',proc:'procedural_pile',ttm:'ttm'};
+const scoreJsonKeys={smol:'smolcua',marc:'marc',proc:'procedural_pile',ttm:'ttm',code:'smolscript'};
 function parseMillions(value){
  if(typeof value==='number')return value/1e6;
  const s=String(value??'').trim().replace(/,/g,'');
@@ -484,7 +484,7 @@ function setupModels(rows){
       evalMeta:isArray?null:(a.evaluation||null),
       checkpointRevision:isArray?null:(a.checkpoint_revision||null)};
    m.millions=parseMillions(m.size);
-   axes.forEach(([k],j)=>m[k]=isArray?a[{smol:3,marc:5,proc:6,ttm:7}[k]]:a.scores[scoreJsonKeys[k]]);
+   axes.forEach(([k],j)=>m[k]=isArray?a[{smol:3,marc:5,proc:6,ttm:7,code:8}[k]]:a.scores[scoreJsonKeys[k]]);
    m.score=axes.every(([k])=>Number.isFinite(m[k]))?axes.reduce((n,[k,,w])=>n+m[k]*w,0):null;
    m.modelCard=modelCardUrl(m.cardRef);
    m.searchIndex=searchIndex(m);
@@ -502,14 +502,15 @@ function validRow(row){return row && typeof row.name==='string' && row.name.trim
    Object.values(scoreJsonKeys).every(k=>row.scores[k]===null||(Number.isFinite(row.scores[k]) && row.scores[k]>=0 && row.scores[k]<=100))}
 // Actual evaluated results are embedded for standalone file:// viewing.
 // On a hosted site data.json remains the authoritative registry.
-const embeddedEvaluatedRows=[{"id":"MinimaLabs/KLM-1.5","name":"KLM-1.5","parameters":"99.0M","model_card_url":"https://huggingface.co/MinimaLabs/KLM-1.5","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A compact language model exploring how much general capability can fit within roughly 100 million parameters.","comment_source":"dotlabs","checkpoint_revision":"b2bbc22327c2b85316c23e871a91b1287aefa795","scores":{"ttm":42.55,"marc":11.5625,"procedural_pile":0.028325667131668543,"smolcua":0.0},"organization":"MinimaLabs","authorLink":"https://huggingface.co/MinimaLabs","authorColor":"#99D9BD"},{"id":"HuggingFaceTB/SmolLM2-135M","name":"SmolLM2-135M","parameters":"135M","model_card_url":"https://huggingface.co/HuggingFaceTB/SmolLM2-135M","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A small open foundation model designed to make language modeling accessible within a tight compute budget.","comment_source":"dotlabs","checkpoint_revision":"93efa2f097d58c2a74874c7e644dbc9b0cee75a2","scores":{"ttm":46.7,"marc":5.3125,"procedural_pile":0.004785877737822812,"smolcua":0.0},"organization":"HuggingFaceTB","authorLink":"https://huggingface.co/HuggingFaceTB","authorColor":"#F5D18C"},{"id":"AxiomicLabs/GPT-X2.5-135M","name":"GPT-X2.5-135M","parameters":"135M","model_card_url":"https://huggingface.co/AxiomicLabs/GPT-X2.5-135M","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"An experimental small-scale foundation model focused on learning useful capabilities with limited parameters.","comment_source":"dotlabs","checkpoint_revision":"53959ad51244b913365c2d66a49a301a83cac59c","scores":{"ttm":43.9,"marc":6.25,"procedural_pile":0.0034782608695652162,"smolcua":0.0},"organization":"AxiomicLabs","authorLink":"https://huggingface.co/AxiomicLabs","authorColor":"#A6A0FF"},{"id":"dotlabs/void.1","name":"void.1","parameters":"90.1M","model_card_url":"https://huggingface.co/dotlabs/void.1","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A byte-level foundation model exploring language and reasoning without a conventional text tokenizer.","comment_source":"dotlabs","checkpoint_revision":"786701ec20edc5f7fee47dfedf4f9bb662ca4bb8","scores":{"ttm":41.65,"marc":1.875,"procedural_pile":0.0426930028759097,"smolcua":0.0},"organization":"DotLabs","authorLink":"https://huggingface.co/dotlabs","authorColor":"#FFFFFF"},{"id":"tiiuae/Falcon-H1-Tiny-90M-Instruct","name":"Falcon-H1-Tiny-90M-Instruct","parameters":"90M","model_card_url":"https://huggingface.co/tiiuae/Falcon-H1-Tiny-90M-Instruct","model_type":"creator_posttrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"An instruction-tuned Falcon-H1 Tiny model from its original creators, aimed at efficient conversational tasks.","comment_source":"dotlabs","checkpoint_revision":"e6389502a0b12cd8da894b395ba5bf7436873b16","scores":{"ttm":43.3,"marc":7.1875,"procedural_pile":7.969346397750256,"smolcua":0.0},"organization":"TII","authorLink":"https://huggingface.co/tiiuae","authorColor":"#D5ACFF"},{"id":"LiquidAI/LFM2.5-230M","name":"LFM2.5-230M","parameters":"230M","model_card_url":"https://huggingface.co/LiquidAI/LFM2.5-230M","model_type":"creator_posttrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A small creator-post-trained Liquid Foundation Model optimized for responsive, efficient text assistance.","comment_source":"dotlabs","checkpoint_revision":"40cb2ad3b3044d5a41eee083a6103c8b523afa45","scores":{"ttm":48.2,"marc":0.0,"procedural_pile":2.780199447600087,"smolcua":8.1778},"organization":"LiquidAI","authorLink":"https://huggingface.co/LiquidAI","authorColor":"#9AB5FF"},{"id":"LiquidAI/LFM2.5-350M","name":"LFM2.5-350M","parameters":"350M","model_card_url":"https://huggingface.co/LiquidAI/LFM2.5-350M","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A lightweight Liquid Foundation Model focused on delivering practical capability with a small deployment footprint.","comment_source":"dotlabs","checkpoint_revision":"9921cd4d82c0e3986887aa4b376f9446729ef8a3","scores":{"ttm":46.85,"marc":0.0,"procedural_pile":11.993131773709633,"smolcua":14.0053},"organization":"LiquidAI","authorLink":"https://huggingface.co/LiquidAI","authorColor":"#9AB5FF"},{"id":"google/gemma-3-270m","name":"gemma-3-270m","parameters":"270M","model_card_url":"https://huggingface.co/google/gemma-3-270m","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A compact pretrained Gemma 3 foundation model intended for efficient, specialized downstream applications.","comment_source":"dotlabs","checkpoint_revision":"9b0cfec892e2bc2afd938c98eabe4e4a7b1e0ca1","scores":{"ttm":46.1,"marc":2.8125,"procedural_pile":0.0029742541128357264,"smolcua":0.0},"organization":"Google","authorLink":"https://huggingface.co/google","authorColor":"#E8B77B"},{"id":"ibm-granite/granite-4.0-350m","name":"granite-4.0-350m","parameters":"350M","model_card_url":"https://huggingface.co/ibm-granite/granite-4.0-350m","model_type":"creator_posttrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"IBM’s compact, creator-post-trained Granite model for practical language, code, and tool-use tasks.","comment_source":"dotlabs","checkpoint_revision":"bd8a1497065c0d6ba1ef19af6b0d2b14bacf71c2","scores":{"ttm":52.7,"marc":2.5,"procedural_pile":13.301300628673525,"smolcua":13.8532},"organization":"IBM Granite","authorLink":"https://huggingface.co/ibm-granite","authorColor":"#9DB5EB"},{"id":"ibm-granite/granite-4.0-h-350m","name":"granite-4.0-h-350m","parameters":"340M","model_card_url":"https://huggingface.co/ibm-granite/granite-4.0-h-350m","model_type":"creator_posttrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"IBM’s hybrid attention–Mamba2 Granite model, post-trained for efficient instruction and tool-oriented tasks.","comment_source":"dotlabs","checkpoint_revision":"3b17b717b8f2f5d305b0a92c1491e239aeda19c8","scores":{"ttm":51.4,"marc":3.125,"procedural_pile":15.574583666404257,"smolcua":15.9499},"organization":"IBM Granite","authorLink":"https://huggingface.co/ibm-granite","authorColor":"#9DB5EB"},{"id":"GODELEV/Rose-1.5-Medium","name":"Rose-1.5-Medium","parameters":"98.6M","model_card_url":"https://huggingface.co/GODELEV/Rose-1.5-Medium","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-05","date_added":"2026-11-05","author_note":"A ~98M-parameter Rose X1.5 model exploring grouped-query attention, refresh gates, and efficient learning within a tight parameter budget.","comment_source":"dotlabs","checkpoint_revision":"02f05fa792cda22ca38b20babe6418da0db02cd9","scores":{"ttm":43.55,"marc":3.4375,"procedural_pile":0.012771832822402176,"smolcua":0},"organization":"GODELEV","authorLink":"https://huggingface.co/GODELEV","authorColor":"#F0A4D5"},{"id":"GODELEV/Rose-1.5-Pro","name":"Rose-1.5-Pro","parameters":"148.8M","model_card_url":"https://huggingface.co/GODELEV/Rose-1.5-Pro","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-05","date_added":"2026-11-05","author_note":"A larger Rose X1.5 experiment using grouped-query attention and refresh gates to push compact language-model capability.","comment_source":"dotlabs","checkpoint_revision":"96bf23ef5d9a468450b60bcc7e5a828ff3102036","scores":{"ttm":45.6,"marc":6.875,"procedural_pile":0.027063257654258682,"smolcua":0},"organization":"GODELEV","authorLink":"https://huggingface.co/GODELEV","authorColor":"#F0A4D5"},{"id":"fromziro/50M-pro","name":"50M-pro","parameters":"49.6M","model_card_url":"https://huggingface.co/fromziro/50M-pro","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-05","date_added":"2026-11-05","author_note":"A compact linear-attention language model built around Gated DeltaNet, exploring efficient text generation at roughly 50 million parameters.","comment_source":"dotlabs","checkpoint_revision":"0da1f2c029bf9b4c0aabfbeab6bd81090c570da1","scores":{"ttm":40.2,"marc":4.0625,"procedural_pile":0.004968944099378885,"smolcua":0},"organization":"FromZero","authorLink":"https://huggingface.co/fromziro","authorColor":"#A4D4B1"},{"id":"bench-labs/tinctura-v1","name":"tinctura-v1","parameters":"96.2M","model_card_url":"https://huggingface.co/bench-labs/tinctura-v1","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-05","date_added":"2026-11-05","author_note":"A from-scratch compact language model using a smaller Cagliostro-style architecture and a 75-billion-token training run.","comment_source":"dotlabs","checkpoint_revision":"f475b8a2698b0f2de14fbae584014908fec4f719","scores":{"ttm":44.6,"marc":10.625,"procedural_pile":0.019425921663509163,"smolcua":0},"organization":"Bench Labs","authorLink":"https://huggingface.co/bench-labs","authorColor":"#EFB48C"},{"id":"bench-labs/cagliostro-v3.5","name":"cagliostro-v3.5","parameters":"146M","model_card_url":"https://huggingface.co/bench-labs/cagliostro-v3.5","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-05","date_added":"2026-11-05","author_note":"An extension of Cagliostro-v3 with a brief additional training stage, exploring stronger capability at the same 146M-parameter scale.","comment_source":"dotlabs","checkpoint_revision":"2e3a2cdb72c0e979d498f2db00ed599e016ac95a","scores":{"ttm":46.55,"marc":7.1875,"procedural_pile":0.007577656444398657,"smolcua":0},"organization":"Bench Labs","authorLink":"https://huggingface.co/bench-labs","authorColor":"#EFB48C"}];
+const embeddedEvaluatedRows=[{"id":"MinimaLabs/KLM-1.5","name":"KLM-1.5","parameters":"99.0M","model_card_url":"https://huggingface.co/MinimaLabs/KLM-1.5","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A compact language model exploring how much general capability can fit within roughly 100 million parameters.","comment_source":"dotlabs","checkpoint_revision":"b2bbc22327c2b85316c23e871a91b1287aefa795","scores":{"ttm":42.55,"marc":11.5625,"procedural_pile":0.028325667131668543,"smolcua":0.0,"smolscript":0.0},"cognirank_index":7.814910666764562,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"d9d2d7c6abb47f86fed6c87cf95e189f4686a7705fc8f3f3cb16b7b6dcc0958a","smolscript":"816dda142e9c336adace796e59bc04e139c0472ee729c12d50edf67913ac8032"}},"organization":"MinimaLabs","authorLink":"https://huggingface.co/MinimaLabs","authorColor":"#99D9BD"},{"id":"HuggingFaceTB/SmolLM2-135M","name":"SmolLM2-135M","parameters":"135M","model_card_url":"https://huggingface.co/HuggingFaceTB/SmolLM2-135M","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A small open foundation model designed to make language modeling accessible within a tight compute budget.","comment_source":"dotlabs","checkpoint_revision":"93efa2f097d58c2a74874c7e644dbc9b0cee75a2","scores":{"ttm":46.7,"marc":5.3125,"procedural_pile":0.004785877737822812,"smolcua":0.0,"smolscript":0.5374255374255374},"cognirank_index":7.125861081745702,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"b863c457cb2b5d6f32650f8a55b0cf90df150391aca002c923eef1473ceb054d","smolscript":"b4e8f5fa96434823307edee7e14b66fe8b9fc9be048ebcf7c00800ae4474270a"}},"organization":"HuggingFaceTB","authorLink":"https://huggingface.co/HuggingFaceTB","authorColor":"#F5D18C"},{"id":"AxiomicLabs/GPT-X2.5-135M","name":"GPT-X2.5-135M","parameters":"135M","model_card_url":"https://huggingface.co/AxiomicLabs/GPT-X2.5-135M","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"An experimental small-scale foundation model focused on learning useful capabilities with limited parameters.","comment_source":"dotlabs","checkpoint_revision":"53959ad51244b913365c2d66a49a301a83cac59c","scores":{"ttm":43.9,"marc":6.25,"procedural_pile":0.0034782608695652162,"smolcua":0.0,"smolscript":0.2849002849002849},"cognirank_index":6.918764954005228,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"89c8ea23447a276d4394e157f7b4bd58778545eb1d6b84c8759b161f4379cdb5","smolscript":"acfee735d0491c330985a3a1b38a75281815f99be4418a9e0acf9e97986aecac"}},"organization":"AxiomicLabs","authorLink":"https://huggingface.co/AxiomicLabs","authorColor":"#A6A0FF"},{"id":"dotlabs/void.1","name":"void.1","parameters":"90.1M","model_card_url":"https://huggingface.co/dotlabs/void.1","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A byte-level foundation model exploring language and reasoning without a conventional text tokenizer.","comment_source":"dotlabs","checkpoint_revision":"786701ec20edc5f7fee47dfedf4f9bb662ca4bb8","scores":{"ttm":41.65,"marc":1.875,"procedural_pile":0.0426930028759097,"smolcua":0.0,"smolscript":0.0},"cognirank_index":5.664777474289665,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"4a79299d13bba736d8f82251d7ab77044af3513b3a5ccdbdf601a89bbdde4926","smolscript":"eb0e2a6253c4158e035e7da9598252f2ac43ed61355e012472ced6d5c8ae2470"}},"organization":"DotLabs","authorLink":"https://huggingface.co/dotlabs","authorColor":"#FFFFFF"},{"id":"tiiuae/Falcon-H1-Tiny-90M-Instruct","name":"Falcon-H1-Tiny-90M-Instruct","parameters":"90M","model_card_url":"https://huggingface.co/tiiuae/Falcon-H1-Tiny-90M-Instruct","model_type":"creator_posttrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"An instruction-tuned Falcon-H1 Tiny model from its original creators, aimed at efficient conversational tasks.","comment_source":"dotlabs","checkpoint_revision":"e6389502a0b12cd8da894b395ba5bf7436873b16","scores":{"ttm":43.3,"marc":7.1875,"procedural_pile":7.969346397750256,"smolcua":0.0,"smolscript":0.8223258223258223},"cognirank_index":8.824853879781008,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"d6a2bdc3501fb7e56855aff8e79a117b33e5b57aec39b814ed97f65f1e6ef85b","smolscript":"570af0a153d49e39125d61cee1cea5368e2425d6d6aad87306c3a7f7a0c6f1fb"}},"organization":"TII","authorLink":"https://huggingface.co/tiiuae","authorColor":"#D5ACFF"},{"id":"LiquidAI/LFM2.5-230M","name":"LFM2.5-230M","parameters":"230M","model_card_url":"https://huggingface.co/LiquidAI/LFM2.5-230M","model_type":"creator_posttrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A small creator-post-trained Liquid Foundation Model optimized for responsive, efficient text assistance.","comment_source":"dotlabs","checkpoint_revision":"40cb2ad3b3044d5a41eee083a6103c8b523afa45","scores":{"ttm":48.2,"marc":0.0,"procedural_pile":2.780199447600087,"smolcua":8.1778,"smolscript":0.0},"cognirank_index":8.739696725810544,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"85f3542c267a43e3ee0eb7af66e2f4ce046b4d25e6f15ee2d2acbd9820f5868b","smolscript":"8e6040d0e2c4e9b6e8c4af8fe9fda4229250cbfb4b0ef9c62b38edd916e3c749"}},"organization":"LiquidAI","authorLink":"https://huggingface.co/LiquidAI","authorColor":"#9AB5FF"},{"id":"LiquidAI/LFM2.5-350M","name":"LFM2.5-350M","parameters":"350M","model_card_url":"https://huggingface.co/LiquidAI/LFM2.5-350M","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A lightweight Liquid Foundation Model focused on delivering practical capability with a small deployment footprint.","comment_source":"dotlabs","checkpoint_revision":"9921cd4d82c0e3986887aa4b376f9446729ef8a3","scores":{"ttm":46.85,"marc":0.0,"procedural_pile":11.993131773709633,"smolcua":14.0053,"smolscript":0.0},"cognirank_index":11.98094563657045,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"23cf0691a4f482164e83c77ca277ba4102c8ea85bd3db222bf3f2f6e9ec7fdce","smolscript":"3a9b27124c8417e6217d3d881bd3aa683a940c9b0b68092b2e03e52a471681fa"}},"organization":"LiquidAI","authorLink":"https://huggingface.co/LiquidAI","authorColor":"#9AB5FF"},{"id":"google/gemma-3-270m","name":"gemma-3-270m","parameters":"270M","model_card_url":"https://huggingface.co/google/gemma-3-270m","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"A compact pretrained Gemma 3 foundation model intended for efficient, specialized downstream applications.","comment_source":"dotlabs","checkpoint_revision":"9b0cfec892e2bc2afd938c98eabe4e4a7b1e0ca1","scores":{"ttm":46.1,"marc":2.8125,"procedural_pile":0.0029742541128357264,"smolcua":0.0,"smolscript":0.0},"cognirank_index":6.415889316655334,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"6e94f6f570384bb6e5af5e52079a1016e66041d5d162e492b255451200ed8590","smolscript":"b8f878ea66e57eed40b3bd22760aeff04d6b2ed26aaebd252043da516f2f21be"}},"organization":"Google","authorLink":"https://huggingface.co/google","authorColor":"#E8B77B"},{"id":"ibm-granite/granite-4.0-350m","name":"granite-4.0-350m","parameters":"350M","model_card_url":"https://huggingface.co/ibm-granite/granite-4.0-350m","model_type":"creator_posttrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"IBM\u2019s compact, creator-post-trained Granite model for practical language, code, and tool-use tasks.","comment_source":"dotlabs","checkpoint_revision":"bd8a1497065c0d6ba1ef19af6b0d2b14bacf71c2","scores":{"ttm":52.7,"marc":2.5,"procedural_pile":13.301300628673525,"smolcua":13.8532,"smolscript":12.960095460095461},"cognirank_index":16.075206592792465,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"2a230917ac62150d383e5641c66324138679c14ed29d323747951f7d1cfa6874","smolscript":"56fd716f2cf690e2a7fd723cbabae9dd783a2c85b8bbd2b482c67cec8375d8dd"}},"organization":"IBM Granite","authorLink":"https://huggingface.co/ibm-granite","authorColor":"#9DB5EB"},{"id":"ibm-granite/granite-4.0-h-350m","name":"granite-4.0-h-350m","parameters":"340M","model_card_url":"https://huggingface.co/ibm-granite/granite-4.0-h-350m","model_type":"creator_posttrained","evaluation_status":"evaluated","date_submitted":"2026-11-04","date_added":"2026-11-04","author_note":"IBM\u2019s hybrid attention\u2013Mamba2 Granite model, post-trained for efficient instruction and tool-oriented tasks.","comment_source":"dotlabs","checkpoint_revision":"3b17b717b8f2f5d305b0a92c1491e239aeda19c8","scores":{"ttm":51.4,"marc":3.125,"procedural_pile":15.574583666404257,"smolcua":15.9499,"smolscript":13.165815665815666},"cognirank_index":17.09199758924824,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"055a122c6216e54ee46d138f3e4e8404a5b8403fed73239bd14e8467b6f20698","smolscript":"439c60e134c754cfa18b1f4a9b04deab4021a2d29a24baaeebd20fb476a519d8"}},"organization":"IBM Granite","authorLink":"https://huggingface.co/ibm-granite","authorColor":"#9DB5EB"},{"id":"GODELEV/Rose-1.5-Medium","name":"Rose-1.5-Medium","parameters":"98.6M","model_card_url":"https://huggingface.co/GODELEV/Rose-1.5-Medium","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-05","date_added":"2026-11-05","author_note":"A ~98M-parameter Rose X1.5 model exploring grouped-query attention, refresh gates, and efficient learning within a tight parameter budget.","comment_source":"dotlabs","checkpoint_revision":"02f05fa792cda22ca38b20babe6418da0db02cd9","scores":{"ttm":43.55,"marc":3.4375,"procedural_pile":0.012771832822402176,"smolcua":0,"smolscript":0.0},"cognirank_index":6.227425649015242,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"3966ed32879da5f22c99cd5b39c4536917af385110688bb5ee4ca1651e4d3d91","smolscript":"a5fa7426a9801dde80cffc21bed32cd5caa27e23e2ea66aefb30bdd6b10e065f"}},"organization":"GODELEV","authorLink":"https://huggingface.co/GODELEV","authorColor":"#F0A4D5"},{"id":"GODELEV/Rose-1.5-Pro","name":"Rose-1.5-Pro","parameters":"148.8M","model_card_url":"https://huggingface.co/GODELEV/Rose-1.5-Pro","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-05","date_added":"2026-11-05","author_note":"A larger Rose X1.5 experiment using grouped-query attention and refresh gates to push compact language-model capability.","comment_source":"dotlabs","checkpoint_revision":"96bf23ef5d9a468450b60bcc7e5a828ff3102036","scores":{"ttm":45.6,"marc":6.875,"procedural_pile":0.027063257654258682,"smolcua":0,"smolscript":0.854885854885855},"cognirank_index":7.384043119957015,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"b303d6e3b8599a2daf7c5cce2f4302b47def18ffd0ac340292c083aea477c28a","smolscript":"098282cdd78b1230f5d882d213c51eeb1cfd80b22d431a3bedf06378856ac2a9"}},"organization":"GODELEV","authorLink":"https://huggingface.co/GODELEV","authorColor":"#F0A4D5"},{"id":"fromziro/50M-pro","name":"50M-pro","parameters":"49.6M","model_card_url":"https://huggingface.co/fromziro/50M-pro","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-05","date_added":"2026-11-05","author_note":"A compact linear-attention language model built around Gated DeltaNet, exploring efficient text generation at roughly 50 million parameters.","comment_source":"dotlabs","checkpoint_revision":"0da1f2c029bf9b4c0aabfbeab6bd81090c570da1","scores":{"ttm":40.2,"marc":4.0625,"procedural_pile":0.004968944099378885,"smolcua":0,"smolscript":0.25252525252525254},"cognirank_index":5.9847090387365,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"85fc0e5e1a94c38ee36c0ced342248c8bfe4be30cea1a0b9b4d6af1855361247","smolscript":"ef0952687e5fe1268f811433da950c461dc57480fc23f9bbb71debaaca830829"}},"organization":"FromZero","authorLink":"https://huggingface.co/fromziro","authorColor":"#A4D4B1"},{"id":"bench-labs/tinctura-v1","name":"tinctura-v1","parameters":"96.2M","model_card_url":"https://huggingface.co/bench-labs/tinctura-v1","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-05","date_added":"2026-11-05","author_note":"A from-scratch compact language model using a smaller Cagliostro-style architecture and a 75-billion-token training run.","comment_source":"dotlabs","checkpoint_revision":"f475b8a2698b0f2de14fbae584014908fec4f719","scores":{"ttm":44.6,"marc":10.625,"procedural_pile":0.019425921663509163,"smolcua":0,"smolscript":0.2849002849002849},"cognirank_index":7.931596040488165,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"006a655293b27ae8efe5576db5cb8a03aed1f45e83d4031fe8be42c4114e600b","smolscript":"97eada25f949017706f31f8d48c23a98fd3a1b2172fbe787ebe8dd3ce9a4c145"}},"organization":"Bench Labs","authorLink":"https://huggingface.co/bench-labs","authorColor":"#EFB48C"},{"id":"bench-labs/cagliostro-v3.5","name":"cagliostro-v3.5","parameters":"146M","model_card_url":"https://huggingface.co/bench-labs/cagliostro-v3.5","model_type":"pretrained","evaluation_status":"evaluated","date_submitted":"2026-11-05","date_added":"2026-11-05","author_note":"An extension of Cagliostro-v3 with a brief additional training stage, exploring stronger capability at the same 146M-parameter scale.","comment_source":"dotlabs","checkpoint_revision":"2e3a2cdb72c0e979d498f2db00ed599e016ac95a","scores":{"ttm":46.55,"marc":7.1875,"procedural_pile":0.007577656444398657,"smolcua":0,"smolscript":0.0},"cognirank_index":7.394753190830399,"evaluation":{"protocol":"cognirank-2026-11-update","requested_context":4096,"component_protocols":{"reasoning":"cognirank-v9","smolscript":"smolscript-0.5.0"},"source_result_sha256":{"reasoning":"dfd91f864c36ca118a014a240e4a0fe80d79b318930d84399f9c1459a5bd5caf","smolscript":"23f98009db8156346cd763c518bf4ed9d821575bc956b88471ea219cea04d51b"}},"organization":"Bench Labs","authorLink":"https://huggingface.co/bench-labs","authorColor":"#EFB48C"}];
 function showEmbeddedEvaluated(){
  setupModels(embeddedEvaluatedRows);
  document.getElementById('app').dataset.source='evaluated';
 }
 const SPACE_DATA_URL='https://huggingface.co/spaces/dotlabs/cognirank/resolve/main/data.json';
 function registryRows(data){
- if(!data||!Array.isArray(data.authors))throw new Error('Invalid data.json: authors[] expected');
+ if(!data||!Array.isArray(data.authors)||data.evaluation_protocol!=='cognirank-2026-11-update')throw new Error('Invalid or outdated CogniRank registry');
+ if(!axes.every(([k,,w])=>Math.abs(data.evaluation_weights?.[scoreJsonKeys[k]]-w)<1e-12))throw new Error('Registry weight mismatch');
  return data.authors.flatMap(author=>{
   if(!author||typeof author.name!=='string'||!Array.isArray(author.models))return [];
   const authorColor=/^#[0-9a-fA-F]{6}$/.test(author.color||'')?author.color:'#FFFFFF';
@@ -635,9 +636,9 @@ function paintRows(animate=false){
  if(!animate&&signature===visibleSignature)return;
  const before=animate?new Map([...tbody.querySelectorAll('tr[data-id]')].map(tr=>[tr.dataset.id,tr.getBoundingClientRect().top])):null;
  const above=begin*height,below=Math.max(0,(filteredRows.length-end)*height);
- tbody.innerHTML=(above?`<tr class="virtual-space" aria-hidden="true"><td colspan="7" style="height:${above}px"></td></tr>`:'')+
+ tbody.innerHTML=(above?`<tr class="virtual-space" aria-hidden="true"><td colspan="8" style="height:${above}px"></td></tr>`:'')+
     filteredRows.slice(begin,end).map((m,i)=>rowMarkup(m,begin+i)).join('')+
-    (below?`<tr class="virtual-space" aria-hidden="true"><td colspan="7" style="height:${below}px"></td></tr>`:'');
+    (below?`<tr class="virtual-space" aria-hidden="true"><td colspan="8" style="height:${below}px"></td></tr>`:'');
  visibleSignature=signature;
  if(animate&&!prefersReducedMotion.matches){
    for(const tr of tbody.querySelectorAll('tr[data-id]')){
@@ -676,7 +677,7 @@ function radarPath(values,mask){
 
 const svg=(tag,attrs)=>{let el=document.createElementNS(NS,tag);Object.entries(attrs||{}).forEach(([k,v])=>el.setAttribute(k,v));return el;};
 for(let v=20;v<=100;v+=20)base.appendChild(svg('polygon',{points:pos(Array(axes.length).fill(v)),class:'radar-ring'}));
-const taskNames={smol:'SmolCUA',marc:'MARC',proc:'Procedural Pile',ttm:'TTM'};
+const taskNames={smol:'SmolCUA',marc:'MARC',proc:'Procedural Pile',ttm:'TTM',code:'SmolScript'};
 // A second plot shares the same SVG viewport; swapping views never shifts the page.
 const barsBase=document.getElementById('bar-base'),barsData=document.getElementById('bar-data');
 // Compare readability: texture is assigned only when selected colors look alike.
@@ -948,7 +949,7 @@ function toggleChartStyle(){
  radarView.setAttribute('aria-hidden',String(toBars));
  chartToggle.setAttribute('aria-pressed',String(toBars));
  // Keep chart switching undisclosed in the visible UI (a small easter egg).
- document.getElementById('radar').setAttribute('aria-label',toBars?'Vertical bars showing scores for selected models':'Four-axis radar showing scores for selected models');
+ document.getElementById('radar').setAttribute('aria-label',toBars?'Vertical bars showing scores for selected models':'Five-axis radar showing scores for selected models');
 }
 // No permanent legend or labels are added: tooltip is revealed only near a bar or vertex.
 const inspect=document.getElementById('chart-inspect');