GitHub 新星
10 条
arXiv 论文
10 条
-
1
Learning to Stop without Learning to Stop: Self-Supervised Confidence Training Improves Reasoning Efficiency
Reasoning models often generate very long reasoning traces, making inference computationally expensive. Existing approaches typically improve efficiency either through inference-time early-stopping meParsa Hosseini, Akasha Tigalappanavara, Sumit Nawathe
-
2
Gap-free Differentially Private PCA for Gaussian Data
We give a gap-free differentially private algorithm for the principal component analysis (PCA) problem with Gaussian data.Alina Ene, Huy L. Nguyen
-
3
First-Order Stationarity of Reverse Diffusions
Recent literature has shown a strong connection between optimization and sampling. We develop the corresponding first-order theory for diffusion models. First, the SDE-based reverse-time flows of overZhifeng Chen, Chenyang Jiang, Yazhen Wang
-
4
Statistical attribute alignment for black-box generative AI via output post-processing
Generative AI systems are increasingly used, but aligning their outputs with user requirements poses a continuing challenge. Here, we aim to ensure that the distribution of an attribute of an AI-generKevin Jiang, Morgane Austern, Edgar Dobriban
-
5
User Model Extraction via Belief Self-Distillation
Large language models (LLMs) implicitly infer attributes of their users and adapt their behavior accordingly, yet these beliefs remain difficult to inspect and causally manipulate. We introduce BeliefAli Holmov, Yiran Huang, Kirill Bykov
-
6
New LoRA Skills Should Read but Never Write
Low-rank adapters (LoRA) make it cheap to fine-tune a large language model once per task, but combining several independently trained adapters into one model remains difficult: merging the updates inZeyan Li, Panqi Yang, Qirong Guo
-
7
Common-Mode Collapse and Recovery in Direct Feedback Alignment
Direct feedback alignment (DFA) trains hidden layers through fixed random projections of output error. With tanh hidden units and independent sigmoid outputs, plain stochastic gradient descent can staVarun Reddy, Bernardo L. Sabatini, Houman Safaai
-
8
Compact Documentation for Coding Agents: A Benchmark, an Optimizer, and Why It Does Not Transfer
We investigate whether natural-language documentation helps coding agents resolve software issues, and we build the tools to construct and evaluate it. We introduce a roundtrip benchmark that scores cMd Shohel Arman, Igor Molybog
-
9
Trust Guided Decision Transformer
Decision Transformer performance degrades on long rollouts because the conditioning context drifts out of the training distribution. We show that this drift is visible through the model's own next staChainesh Gautam, Raghuram Bharadwaj Diddigi, Chandramouli Kamanchi
-
10
OC-GS: Gaussian Splatting for Irregular Turntable Capture
Uneven rotation and dropped frames make equal-angle assumptions unreliable for turntable reconstruction. We present OC-GS, an object-centric Gaussian splatting that refines each image's angle while maJae Joong Lee, Bedrich Benes
Hacker News
10 条
- 1 Sonnet 5.5
- 2 Pirating the Pirates
- 3 Parley: Federated, decentralised chat that speaks plain IRC
- 4 Kids turned low-traffic NPR Spotify comments into a secret group chat
- 5 It's Time to Investigate the AI Labs
- 6 Jeff – Jev-compatible 0.8B decision models, trained at home, ~30 ms
- 7 Hijacking the PS5's RTMP stream
- 8 Updated Google Maps shows destruction of the city of Rafah
- 9 World Labs Is Joining AMD
- 10 Show HN: HN.watch – Videos of all Hacker News posts