Showing papers from Technische Universität Berlin Show all papers
Atlanta Poster Session 1, Wed, Dec 9, 10:00 AM–1:00 PM, Hall C1 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
Byzantine agreement with predictors achieves tight consistency-robustness trade-offs and linear resilience degradation with prediction errors.
Paris Poster Session 5, Fri, Dec 11, 11:30 AM–1:30 PM, Paris Poster Hall · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
MASEM samples disconnected manifolds by entropy maximization via resampling, exponentially reducing KL divergence and improving Sinkhorn distance by an order of magnitude.
Atlanta Poster Session 5, Fri, Dec 11, 10:00 AM–1:00 PM, Hall C1 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
Symb-xMIL quantifies alignment between MIL predictions and human-readable logical rules to expose decision patterns, recover ground-truth rules, and refine survival stratification beyond HPV status.
Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
Frank-Wolfe achieves Ω(T^{-p/(p-1)}) lower bounds on p-uniformly convex sets for p ≥ 3 under exact line search or short steps, matching upper bounds via low-dimensional dynamics.
Paris Poster Session 4, Thu, Dec 10, 5:30 PM–7:30 PM, Paris Poster Hall · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
SecureClaw dual-bounds LLM agents by confining plaintext via opaque handles at the read boundary and enforcing authorized previews at the action sink, achieving near-zero attack success with preserved utility.
Paris Poster Session 6, Fri, Dec 11, 2:30 PM–4:30 PM, Paris Poster Hall · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
LLM inference components cause detectable output deviations enabling fingerprinting of engines, attention backends, and hardware from queries.
Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
Vision-language models use celebrity identity shortcuts rather than visual age cues, and activation steering suppresses this to cut mean absolute error by up to 25%.
Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
Learning-augmented online scheduling achieves O(1)-competitive latency with O(1) preemptions per job on parallel machines, with overhead scaling logarithmically in prediction error.
Atlanta Poster Session 5, Fri, Dec 11, 10:00 AM–1:00 PM, Hall C1 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
Structural causal bottleneck models assume causal effects depend on low-dimensional cause summaries, enabling flexible dimension reduction via standard algorithms, improved low-sample transfer, and identifiable bottlenecks.
Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.