Linear attention keeps a state of fixed size. At long context, many stored items share this state, and interference between them degrades retrieval. Gated DeltaNet-2 (GDN-2), like every delta-rule model before it, derives its erase vector from the key of the current token. However, the interference in its reads is meas...
This work formalizes the two levels of active sequential hypothesis testing as an outer probe-design problem and an inner identification problem, and shows that selecting which evaluations to construct at minimum cost, so that every pair of candidates is distinguished, is exactly a weighted set cover problem.
This work studies the LLC of a graph attention model by exploiting symmetries in both the graph structure and the attention parameters, and develops an analytic framework through a teacher--student setting and explicit LLC estimates after considering the symmetry--induced degeneracies.
Vishnu Varadarajan, Mihir More, Aritra Das et al.· 0 citations
This work studies how activation-based detection changes when agents are explicitly informed that their internal activations are being monitored, and when they additionally receive feedback from the monitor.
Aritra Das, Jaee Ponde, Mihir More et al.· 0 citations
These results resolve the corresponding questions regarding the parameterized complexity status for zonotope norm maximization and two-layer ICNN Lipschitz constants.
Aritra Das, Vincent Froese, Moritz Grillo et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
