GitHub 新星
10 条
arXiv 论文
10 条
-
1
Objective vs. Search: Decomposing What Makes a Good Tokeniser
Two dominant tokenisation algorithms are used by modern language models: byte-pair encoding (BPE) and UnigramLM. These differ along two orthogonal axes: their optimisation objective (compression vs. lAhmetcan Yavuz, Clara Meister, Tiago Pimentel
-
2
A Zeroth-Order Paradigm for LLM Preference Alignment
Direct preference alignment methods are widely used to align large language models (LLMs) with human preferences because of their computational and memory efficiency. However, likelihood displacementPeter Chen, Xi Chen, Wotao Yin
-
3
Dreaming the Sound of Contact: Leveraging Video and Audio Generation for Zero-Shot Force-Aware Manipulation and Data Generation
Recent advances in video generation allow robots to learn manipulation trajectories from generated videos. However, these approaches produce purely kinematic trajectories that lack force information,Guanhua Ji, Tianyu Li, Dayoon Suh
-
4
Exponential Hardness of Off-Policy Evaluation under History-Dependent Logging
Can a logged dataset visit every hidden state frequently and still be exponentially uninformative about a target policy's value? We show that it can when the logger depends on history. For every horizPranaya Jajoo
-
5
Cognitive Extensions for Dual-Process Language Agents: Memory and Self-Reflection in Interactive Environments
Language agents remain brittle in interactive environments, where success requires long-horizon state tracking, valid action execution, and recovery from failed steps. We extend SwiftSage, a dual-procJoão Meneses dos Santos, Arlindo L. Oliveira
-
6
Affora: A Design System for Agent-Friendly Interfaces
Computer-use agents increasingly operate software designed for people, but interfaces often leave actions or task state unclear to machine readers. We present Affora, a design system that supports botJin Gao
-
7
Flag Game: A Toy Model for Mechanistic Swarm Interpretability
Emergent coordinated behaviors of AI agents are starting to present critical safety risks. A key phenomenon driving these behaviors is the rapid formation and spread of beliefs about the world, and meElizabeth Pavlova, Hidenori Tanaka
-
8
How Model Growth, Recursion, and Boundary Operators Influence Scaling Exponents
Scaling laws predict how loss decreases with increases in computation. We show, contrary to conventional wisdom, that architectural interventions can modify scaling exponents in pre-training, leadingZixi Chen, Akshay Vegesna, Samip Dahal
-
9
rMuscle: Robotic Muscle Memory for Efficient Vision-Language-Action Model Inference
Factory work is a promising early scenario for embodied AI: assigning repetitive manual jobs to robots has clear economic payoff, and a structured station keeps the jobs tractable for current policiesKaijun Zhou, Zhiyang Li, Le Chen
-
10
Monitoring and Discovering Reward Hacking with Internal Representations during LLM Evaluations
As models scale, reward hacking becomes more frequent, more sophisticated, and more consequential. Does it leave a telltale signature in model representations? This work analyzes how reward hacking isLeon Bergen, Usha Bhalla, Andrew Lee
Hacker News
10 条
- 1 Fujitsu launches made-in-Japan next-generation CPU FUJITSU-MONAKA
- 2 Hister: A private search engine for the pages you visit and the files you keep
- 3 How GLM built its own inference infrastructure
- 4 CCC invites all model citizens to 40C3
- 5 My temporary PHP fix from 2014 has nearly 20M installs. Today I'm deprecating it
- 6 Astra for Law
- 7 Bend – A language that blocks AI mistakes via proof, on CPU and GPU
- 8 Wax motor
- 9 Why I didn’t sign the Fields medallists’ letter
- 10 The American Religion of Self-Storage Facilities