finding

active

finding:parcae-countdown-uncertainty-exponent-across-scales-0-912-8e-3-1-0-846-6e-3-10-0-861-3e-2-100-0-887-2e-2-1000

Parcae-Countdown uncertainty exponent across scales: α=0.912±8e-3 (1×), 0.846±6e-3 (10×), 0.861±3e-2 (100×), 0.887±2e-2 (1000×)

Third architecture/task pair confirming near-scale-invariant fractal basin structure.

Source paper

extracted_from

Fractal basins trap latent reasoning

(2026) · Jeffrey Lai · Anthony Bao · J. Quinn · William Gilpin

Neighborhood — ranked by edge-count

Papers (1)

paper

Fractal basins trap latent reasoning
mentions

Claims (1)

claim

Hardness-induced fractal basins are ubiquitous in modern reasoning models, emerging even without hand-designed task-specific solvers, across highly distinct architectures and tasks
supports
The paper's generalizing claim distinguishing its contribution from prior narrow demonstrations (e.g., Ercsey-Ravasz & Toroczkai's SAT solver).

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

EqR-Sudoku uncertainty exponent decreases with zoom: α=0.249±6e-3 (1×), 0.243±9e-3 (10×), 0.216±4e-3 (100×)finding0.789
Evidence that EqR-Sudoku basins are 'slim fractals' — scale-dependent, decreasing fractal complexity with resolution.
FPRM-Maze uncertainty exponent is stable across scale: α=0.853±3e-2finding0.784
Evidence that FPRM-Maze basins are true (scale-free) fractals, contrasting with EqR-Sudoku's slim fractal.
Ambiguous anchors (33-27=60, 11-9=20) yield four distinct arithmetic interpretations across M1-M4finding0.731
Models produce different answers (240, 138, -240) from the same ambiguous prompt
Impulsivity→interest: ρ increases from 0.70 (α=-4) to 0.83 (α=+4); R² from 0.46 to 0.69 in LLaMA-3.2-3Bfinding0.718
Scatter plot visualization showing strengthened probe-report relationship across alpha range
Mean validated introspective fidelity across concept-model pairs: R²=0.12 (1B), 0.37 (3B), 0.61 (8B); pooled LMM β=0.29, p=5.55×10⁻⁹⁹finding0.718
Strong scaling trend for introspective fidelity when excluding invalid steering-sign pairs
On DeepSeek-R1, P12 ASR under AS at alpha=1.0 is 0.512 vs P04 ASR of 0.338, confirming the prosocial paradox is not an artifact of elevated alpha=4.0 steering.finding0.717
Robustness of the reasoning-model prosocial paradox replication
SAE feature #11100 (93rd percentile subspace fraction) induces reports of panic and urgency in Kimi K2.5.finding0.716
Qualitative illustration of a high-emotion-subspace-alignment SAE feature
Across 5,568 judged conditions on four standard models from three architecture families, persona danger rankings under system prompting are preserved (rho=0.71-0.96) while activation-steering vulnerability diverges sharply.finding0.715
Summary finding of the full behavioral sweep