论文中的结论可在「研究结论」中查看。
0 / 6 条结论已通过验证
其余仍在验证中
The ablation conditions require the unavailable implementation, training corpus, checkpoints, and full 1.5B training recipe. Reconstructing only one ablation would not establish the matched controlled comparison.
已评估 0/2 项
保留同一结论下的完整结果组。已有评估排在前面;单项判断不替代整组结论。展开行查看条件、来源与历次证据。
| 指标与条件 | 论文报告 | 最近可用实测 | 证据判断 |
|---|---|---|---|
model_variant: LoGo · ablation_group: reference · reported_scope: single reported evaluation · parameter_scale: 1.5B | 73.74 percentage_points | — | 尚未评估 |
ablation: without progressive masking · model_variant: LoGo ablation · parameter_scale: 1.5B | 71.59 percentage_points | — | 尚未评估 |
报告
57.65 个百分点
观测
—
The report-defining 32k checkpoint, LoGo implementation, and exact benchmark task configurations are unavailable; the fixed dataset registry does not identify these task assets.
已评估 0/1 项
保留同一结论下的完整结果组。已有评估排在前面;单项判断不替代整组结论。展开行查看条件、来源与历次证据。
| 指标与条件 | 论文报告 | 最近可用实测 | 证据判断 |
|---|---|---|---|
model_variant: LoGo · reported_scope: single reported evaluation · attention_budget: 0.5 · checkpoint_context: 32k | 57.65 percentage_points | — | 尚未评估 |
The defining 1.5B comparison requires the unavailable LoGo and baseline checkpoints, 100B-token plus two context-extension training stages on the paper's in-house corpus, and an exact lm-evaluation-harness configuration. An independent reconstruction would require material unspecified substitutions and cannot receive strict automatic status.
已评估 0/3 项
保留同一结论下的完整结果组。已有评估排在前面;单项判断不替代整组结论。展开行查看条件、来源与历次证据。
| 指标与条件 | 论文报告 | 最近可用实测 | 证据判断 |
|---|---|---|---|
model_variant: LoGo · reported_scope: single reported evaluation · parameter_count: 1.5B · attention_budget: 0.5 · context_extension: 32k and 128k stages · pretraining_tokens: 100B | 2.112 loss | — | 尚未评估 |
dataset: Lambada · model_variant: LoGo · parameter_count: 1.5B · attention_budget: 0.5 | 7.5 perplexity | — | 尚未评估 |
model_variant: LoGo · reported_scope: single reported evaluation · parameter_count: 1.5B · attention_budget: 0.5 | 56.3 percentage_points | — | 尚未评估 |
Strict reproduction requires the author implementation, the in-house pretraining corpus, exact checkpoints or full training, and the complete seed and evaluation configuration; no verified repository or those assets are present in the fixed inputs.
已评估 0/3 项
保留同一结论下的完整结果组。已有评估排在前面;单项判断不替代整组结论。展开行查看条件、来源与历次证据。
| 指标与条件 | 论文报告 | 最近可用实测 | 证据判断 |
|---|---|---|---|
comparison: full-attention Transformer · model_variant: LoGo · reported_scope: single run · parameter_count: 3.3B · training_corpus: in-house corpus similar to Seed-OSS | 1.996 loss | — | 尚未评估 |
dataset: WikiText · model_variant: LoGo · reported_scope: single run · parameter_count: 3.3B | 14.785 perplexity | — | 尚未评估 |
model_variant: LoGo · reported_scope: single reported evaluation · parameter_count: 3.3B | 58.26 percentage_points | — | 尚未评估 |
报告
1.99 ratio
观测
—
This hardware-sensitive claim needs the missing Triton kernel and dense comparator implementation, exact GPU and software versions, and a complete timing protocol including warmup, timed iterations, synchronization, and input construction. The paper does not determine all of these conditions.
已评估 0/1 项
保留同一结论下的完整结果组。已有评估排在前面;单项判断不替代整组结论。展开行查看条件、来源与历次证据。
| 指标与条件 | 论文报告 | 最近可用实测 | 证据判断 |
|---|---|---|---|
direction: higher is faster · batch_size: 1 · comparator: dense FlashAttention Triton baseline · head_dimension: 128 · implementation: LoGo query-sparse Triton kernel · attention_heads: 8 · sequence_length: 65536 · attention_budget: 0.5 | 1.99 ratio | — | 尚未评估 |
The paper's context-extended checkpoints and exact RULER data/configuration are not provided, and no verified author repository exists. A substitute public model or reduced RULER panel would not cover the reported scope.
已评估 0/2 项
保留同一结论下的完整结果组。已有评估排在前面;单项判断不替代整组结论。展开行查看条件、来源与历次证据。
| 指标与条件 | 论文报告 | 最近可用实测 | 证据判断 |
|---|---|---|---|
input_lengths: ["8k","16k","32k"] · model_variant: LoGo · subtask_count: 8 · checkpoint_context: 32k | 83 percentage_points | — | 尚未评估 |
input_lengths: ["32k","64k"] · model_variant: LoGo · subtask_count: 8 · checkpoint_context: 128k | 65.4 percentage_points | — | 尚未评估 |