Notice: This website is continuously updated as new conference proceedings and journal volumes are published. Consequently, the figures presented here may differ from those reported in our publications.
A brief overview of our methodology is provided on the Methods page, along with references to publications documenting our approach and links to the code and data used in each study.
The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates.
On the convergence of no-regret learning in selfish routing
Authors: Walid Krichene, Benjamin Drighès, Alexandre Bayen
ICML 2014 | Venue PDF | LLM Run Details | Input Tokens: 16,366 Total number of tokens sent to the LLM as input for this paper's analysis. | Output Tokens: 4,728 Total number of tokens produced by the LLM (including reasoning/thinking tokens) for this paper's analysis.
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | We show that convergence holds for a large class of online learning algorithms, inspired from the continuous-time replicator dynamics. In particular, the discounted Hedge algorithm is proved to belong to this class, which guarantees its convergence. and Theorem 2. If for all k, the population strategies (µk(τ))τ satisfy an AREP algorithm with sublinear regret, then the sequence (µ(τ)) converges to the set of Nash equilibria. |
| Researcher Affiliation | Academia | Walid Krichene EMAIL University of California, 652 Sutardja Dai Hall, Berkeley, CA 94720 USA Benjamin Drigh es EMAIL Ecole Polytechnique, Route de Saclay, 91120 Palaiseau, France Alexandre Bayen EMAIL University of California, 642 Sutardja Dai Hall, Berkeley, CA 94720 USA |
| Pseudocode | Yes | Algorithm 1 Online learning setting Input: For every player x Xk, a learning algorithm (hx τ)τ and initial distribution π(0)(x) Pk. 1: for each time step τ do 2: Every player x independently draws a path A(τ)(x) π(τ)(x). 3: For all k, the vector of path losses ℓk(µ(τ)) is revealed to players in Xk. Players incur losses corresponding to their path choice. 4: Every player updates her strategy: π(τ+1)(x) = hx τ (ℓk(µ(t)))t τ, π(τ)(x) . |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described, nor does it include specific repository links or explicit code release statements. |
| Open Datasets | No | Figure 2 mentions that 'Latency functions are taken to be quadratic increasing, and generated randomly', indicating that no publicly available or open dataset with concrete access information was used. |
| Dataset Splits | No | The paper does not provide specific dataset split information, exact percentages, sample counts, or detailed splitting methodology needed to reproduce data partitioning. |
| Hardware Specification | No | The paper does not provide specific hardware details (exact GPU/CPU models, processor types, memory amounts, or detailed computer specifications) used for running its experiments or illustrations. |
| Software Dependencies | No | The paper does not provide specific ancillary software details, such as library or solver names with version numbers, needed to replicate any computational aspects. |
| Experiment Setup | Yes | Figure 2 states 'Latency functions are taken to be quadratic increasing, and generated randomly.' and provides specific learning rates: 'With a constant learning rate γ = 0.7, (µk(τ))τ does not converge (left). With a harmonic sequence of learning rates, γτ = 1 1+τ/10, (µk(τ))τ converges to the set of Nash equilibria.' |