Learn Agentic RL
A brief introduction to Agentic Reinforcement Learning. And some of my thoughts.
A brief introduction to Agentic Reinforcement Learning. And some of my thoughts.
A survey of emerging RL-based approaches to enhance LLM-based decompilation, covering D-LiFT, SK2Decompile, and RlDecompiler.
Paper read of FoxDec, which claims to be able to lift C-compiled x86-64 binaries with provably sound overapproximation using Hoare Graphs verified in Isabelle/HOL.
哥们正在高高兴兴的准备一堆用于进行逆向工程的工具,结果Windows Defender给我把这些玩意全都删了。甚至还警告说这些东西是”hack”;不是哥们,我要的可不就是hack吗! 因此我就想把Windows Defender给关了,但是最后感觉还是太不稳妥了。要是哥们真的不小心下了个神秘病毒把我的主环境干翻了,那可就不妙了。 因此我决定自己整一个Windows虚拟环境,里面放上各种工...
在网上发现了开源的x86-64 CPU模拟器
Analyzing and deobfuscating Polaris Obfuscator's AliasAccess pass, which routes local variable accesses through randomized struct packing and multi-hop pointer chains.
Weakest precondition as a backward predicate transformer — core rules, binary IR implementation with Miasm/PyVEX, and application to patch verification.
Reaching definition analysis in static program analysis — a forward may-analysis determining which definitions may reach each program point without being overwritten.
Live variable analysis in static program analysis — a backward may-analysis determining which variables may be used before redefinition.
A description of forward-flow analysis in static program analysis (SPA).