ICLR 2026 - Submissions
Submissions
Summary Statistics
| Quantity AI Content | Count | Avg Rating |
|---|---|---|
| 0-10% | 0 (0%) | N/A |
| 10-30% | 0 (0%) | N/A |
| 30-50% | 0 (0%) | N/A |
| 50-70% | 1 (100%) | 2.50 |
| 70-90% | 0 (0%) | N/A |
| 90-100% | 0 (0%) | N/A |
| Total | 1 (100%) | 2.50 |
| Title | Abstract | Avg Rating | Quantity AI Content | Reviews | Pangram Dashboard |
|---|---|---|---|---|---|
| Test-Time Layer Recurrence Enables Ultra-Deep Thinking in LLMs Without Chain-of-Thought | Transformers possess a \textbf{neural depth} of only $O(1)$, which restricts them to solving primarily \textbf{inductive} reasoning problems of bounded depth. In contrast, recurrent models allow the l... | 2.50 | 66% | See Reviews | View AI Dashboard |