Ipotesi della Rappresentazione del Frame: Interpretabilità del Modello Linguistico a Lungo Termine con Token Multipli e Generazione di Testo Guidata dal Concetto
Frame Representation Hypothesis: Multi-Token LLM Interpretability and Concept-Guided Text Generation
December 10, 2024
Autori: Pedro H. V. Valois, Lincon S. Souza, Erica K. Shimomoto, Kazuhiro Fukui
cs.AI
Abstract
L'interpretabilità è una sfida chiave nel favorire la fiducia nei Large Language Models (LLM), che deriva dalla complessità dell'estrazione del ragionamento dai parametri del modello. Presentiamo l'Ipotesi della Rappresentazione a Cornice, un quadro teoricamente robusto basato sull'Ipotesi della Rappresentazione Lineare (LRH) per interpretare e controllare i LLM modellando parole multi-token. Ricerche precedenti hanno esplorato la LRH per collegare le rappresentazioni dei LLM a concetti linguistici, ma erano limitate all'analisi di singoli token. Poiché la maggior parte delle parole è composta da diversi token, estendiamo la LRH a parole multi-token, consentendo così l'uso su qualsiasi dato testuale con migliaia di concetti. A tal fine, proponiamo che le parole possano essere interpretate come cornici, sequenze ordinate di vettori che catturano meglio le relazioni tra token e parole. Successivamente, i concetti possono essere rappresentati come la media delle cornici delle parole che condividono un concetto comune. Mostriamo questi strumenti attraverso la Decodifica Guidata da Concetto Top-k, che può guidare in modo intuitivo la generazione di testo utilizzando i concetti scelti. Verifichiamo tali idee sui modelli Llama 3.1, Gemma 2 e Phi 3, dimostrando pregiudizi di genere e linguistici, esponendo contenuti dannosi, ma anche il potenziale per rimediare ad essi, portando a LLM più sicuri e trasparenti. Il codice è disponibile su https://github.com/phvv-me/frame-representation-hypothesis.git
English
Interpretability is a key challenge in fostering trust for Large Language
Models (LLMs), which stems from the complexity of extracting reasoning from
model's parameters. We present the Frame Representation Hypothesis, a
theoretically robust framework grounded in the Linear Representation Hypothesis
(LRH) to interpret and control LLMs by modeling multi-token words. Prior
research explored LRH to connect LLM representations with linguistic concepts,
but was limited to single token analysis. As most words are composed of several
tokens, we extend LRH to multi-token words, thereby enabling usage on any
textual data with thousands of concepts. To this end, we propose words can be
interpreted as frames, ordered sequences of vectors that better capture
token-word relationships. Then, concepts can be represented as the average of
word frames sharing a common concept. We showcase these tools through Top-k
Concept-Guided Decoding, which can intuitively steer text generation using
concepts of choice. We verify said ideas on Llama 3.1, Gemma 2, and Phi 3
families, demonstrating gender and language biases, exposing harmful content,
but also potential to remediate them, leading to safer and more transparent
LLMs. Code is available at
https://github.com/phvv-me/frame-representation-hypothesis.gitSummary
AI-Generated Summary
DeepSeek-R1: Incentivizzare la capacità di ragionamento nei LLM tramite Apprendimento per RinforzoDeepSeek-R1: Incentivizing Reasoning Capability in LLMs via
Reinforcement Learning
DeepSeek-R1: Incentivizzare la capacità di ragionamento nei LLM tramite Apprendimento per Rinforzo
DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via
Reinforcement Learning
DeepSeek-AI, Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, Xiaokang Zhang, Xingkai Yu, Yu Wu, Z. F. Wu, Zhibin Gou, Zhihong Shao, Zhuoshu Li, Ziyi Gao, Aixin Liu, Bing Xue, Bingxuan Wang, Bochao Wu, Bei Feng, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, Damai Dai, Deli Chen, Dongjie Ji, Erhang Li, Fangyun Lin, Fucong Dai, Fuli Luo, Guangbo Hao, Guanting Chen, Guowei Li, H. Zhang, Han Bao, Hanwei Xu, Haocheng Wang, Honghui Ding, Huajian Xin, Huazuo Gao, Hui Qu, Hui Li, Jianzhong Guo, Jiashi Li, Jiawei Wang, Jingchang Chen, Jingyang Yuan, Junjie Qiu, Junlong Li, J. L. Cai, Jiaqi Ni, Jian Liang, Jin Chen, Kai Dong, Kai Hu, Kaige Gao, Kang Guan, Kexin Huang, Kuai Yu, Lean Wang, Lecong Zhang, Liang Zhao, Litong Wang, Liyue Zhang, Lei Xu, Leyi Xia, Mingchuan Zhang, Minghua Zhang, Minghui Tang, Meng Li, Miaojun Wang, Mingming Li, Ning Tian, Panpan Huang, Peng Zhang, Qiancheng Wang, Qinyu Chen, Qiushi Du, Ruiqi Ge, Ruisong Zhang, Ruizhe Pan, Runji Wang, R. J. Chen, R. L. Jin, Ruyi Chen, Shanghao Lu, Shangyan Zhou, Shanhuang Chen, Shengfeng Ye, Shiyu Wang, Shuiping Yu, Shunfeng Zhou, Shuting Pan, S. S. Li, Shuang Zhou, Shaoqing Wu, Shengfeng Ye, Tao Yun, Tian Pei, Tianyu Sun, T. Wang, Wangding Zeng, Wanjia Zhao, Wen Liu, Wenfeng Liang, Wenjun Gao, Wenqin Yu, Wentao Zhang, W. L. Xiao, Wei An, Xiaodong Liu, Xiaohan Wang, Xiaokang Chen, Xiaotao Nie, Xin Cheng, Xin Liu, Xin Xie, Xingchao Liu, Xinyu Yang, Xinyuan Li, Xuecheng Su, Xuheng Lin, X. Q. Li, Xiangyue Jin, Xiaojin Shen, Xiaosha Chen, Xiaowen Sun, Xiaoxiang Wang, Xinnan Song, Xinyi Zhou, Xianzu Wang, Xinxia Shan, Y. K. Li, Y. Q. Wang, Y. X. Wei, Yang Zhang, Yanhong Xu, Yao Li, Yao Zhao, Yaofeng Sun, Yaohui Wang, Yi Yu, Yichao Zhang, Yifan Shi, Yiliang Xiong, Ying He, Yishi Piao, Yisong Wang, Yixuan Tan, Yiyang Ma, Yiyuan Liu, Yongqiang Guo, Yuan Ou, Yuduan Wang, Yue Gong, Yuheng Zou, Yujia He, Yunfan Xiong, Yuxiang Luo, Yuxiang You, Yuxuan Liu, Yuyang Zhou, Y. X. Zhu, Yanhong Xu, Yanping Huang, Yaohui Li, Yi Zheng, Yuchen Zhu, Yunxian Ma, Ying Tang, Yukun Zha, Yuting Yan, Z. Z. Ren, Zehui Ren, Zhangli Sha, Zhe Fu, Zhean Xu, Zhenda Xie, Zhengyan Zhang, Zhewen Hao, Zhicheng Ma, Zhigang Yan, Zhiyu Wu, Zihui Gu, Zijia Zhu, Zijun Liu, Zilin Li, Ziwei Xie, Ziyang Song, Zizheng Pan, Zhen Huang, Zhipeng Xu, Zhongyu Zhang, Zhen Zhang•Jan 22, 2025•3735
Rapporto Tecnico Qwen2.5Qwen2.5 Technical Report
Rapporto Tecnico Qwen2.5
Qwen2.5 Technical Report
Qwen, An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, Junyang Lin, Kai Dang, Keming Lu, Keqin Bao, Kexin Yang, Le Yu, Mei Li, Mingfeng Xue, Pei Zhang, Qin Zhu, Rui Men, Runji Lin, Tianhao Li, Tingyu Xia, Xingzhang Ren, Xuancheng Ren, Yang Fan, Yang Su, Yichang Zhang, Yu Wan, Yuqiong Liu, Zeyu Cui, Zhenru Zhang, Zihan Qiu•Dec 19, 2024•36311
MiniMax-01: Scalare i modelli di base con attenzione lampeggianteMiniMax-01: Scaling Foundation Models with Lightning Attention
MiniMax-01: Scalare i modelli di base con attenzione lampeggiante
MiniMax-01: Scaling Foundation Models with Lightning Attention
MiniMax, Aonian Li, Bangwei Gong, Bo Yang, Boji Shan, Chang Liu, Cheng Zhu, Chunhao Zhang, Congchao Guo, Da Chen, Dong Li, Enwei Jiao, Gengxin Li, Guojun Zhang, Haohai Sun, Houze Dong, Jiadai Zhu, Jiaqi Zhuang, Jiayuan Song, Jin Zhu, Jingtao Han, Jingyang Li, Junbin Xie, Junhao Xu, Junjie Yan, Kaishun Zhang, Kecheng Xiao, Kexi Kang, Le Han, Leyang Wang, Lianfei Yu, Liheng Feng, Lin Zheng, Linbo Chai, Long Xing, Meizhi Ju, Mingyuan Chi, Mozhi Zhang, Peikai Huang, Pengcheng Niu, Pengfei Li, Pengyu Zhao, Qi Yang, Qidi Xu, Qiexiang Wang, Qin Wang, Qiuhui Li, Ruitao Leng, Shengmin Shi, Shuqi Yu, Sichen Li, Songquan Zhu, Tao Huang, Tianrun Liang, Weigao Sun, Weixuan Sun, Weiyu Cheng, Wenkai Li, Xiangjun Song, Xiao Su, Xiaodong Han, Xinjie Zhang, Xinzhu Hou, Xu Min, Xun Zou, Xuyang Shen, Yan Gong, Yingjie Zhu, Yipeng Zhou, Yiran Zhong, Yongyi Hu, Yuanxiang Fan, Yue Yu, Yufeng Yang, Yuhao Li, Yunan Huang, Yunji Li, Yunpeng Huang, Yunzhi Xu, Yuxin Mao, Zehan Li, Zekang Li, Zewei Tao, Zewen Ying, Zhaoyang Cong, Zhen Qin, Zhenhua Fan, Zhihang Yu, Zhuo Jiang, Zijia Wu•Jan 14, 2025•2836