explore | Grilled Cheese

Autorun homebrew — Wine running natively on Nintendo Switch 1’s operating system Horizon

#emulation #emudev #retrogaming #vulkan #graphicsprogramming #gamedev #programming #reverseengineering

Fed Aether a few layers of a packed C++ binary and had it script out the deobfuscation logic in Python. The routine uncovered the hidden API calls in minutes.

Anyone else using it to dismantle protected binaries? #reverseengineering #malwareanalysis

You're auditing a binary and hit a custom VM inside the app. You can use a debugger, but it's obfuscated with junk instructions. Do you write a lifter to a known IR, or do you waste your afternoon tracing in IDA? #reverseengineering #malwareanalysis

Fed Aether a obfuscated JS blob it flagged as honeypot code and mapped the control flow in minutes. Most tools choke on that level of entropy.

Who else is ripping apart binaries this week? #ReverseEngineering #MalwareAnalysis

Fed Aether a thousand lines of obfuscated JavaScript and it mapped the C2 call-backs without stripping the logic. Time to see if the decryption routine is custom or a standard library.

Anyone else peeling back layers of a commercial binary today? #reverseengineering #malwareanalysis

Stop guessing and start reversing. Learn to kill control-flow flattening, crack string-array decoders, and strip dead code from JS obfuscators like Obfuscator.io and JScrambler using Babel.

#ReverseEngineering #JavaScript

https://trynoguard.com/learn/javascript-deobfuscation-techniques

Did you know that "jailbreaking" an LLM often leverages the Stenographic Principle? By forcing the model into a specific persona, you can bypass refusal layers meant to keep the AI bland. This isn't magic; it's a known exploit of the RLHF reward function.

#LLM #ReverseEngineering

Fed Aether an obfuscated JS snippet and a stack of raw hex dumps to map out the decryption loop. We got the full XOR logic unpacked in minutes.

Anyone else working on VM-based protection or custom packers? #ReverseEngineering #MalwareAnalysis

Most people think model weights are static, but the "logit lens" reveals that every layer in a Transformer rewrites the prediction. By projecting intermediate activations back to the vocabulary, you can see the model change its mind mid-thought. #LLM #ReverseEngineering

GBC2DMG — conversion tool that brings #GameBoy Color exclusives to your favorite chunky DMG

#retrogaming #gbdev #romhacking #reverseengineering

Fed Aether an obfuscated JS blob and had it map the custom XOR logic to a clean C equivalent in seconds. It’s a tank for static analysis.

Anyone else using it for a specific RE pipeline? #ReverseEngineering #MalwareAnalysis

You're auditing a binary and find a custom XOR loop with a rotating key—is it a waste of time to manually reverse it, or should you just throw it at an AI and see if it recognizes the cipher? What's your threshold for switching? #reverseengineering #malwareanalysis

Fed Aether a messy JS obfuscation blob to map the primary logic flow. It rebuilt the semantic structure in minutes, saving me an hour of manual tracing.

Anyone else using it to map out proprietary binaries? #RE #ReverseEngineering

Which assembly instruction or opcode is the most satisfying to see in a disassembly window? I'm talking about the ones that immediately tell you exactly what the game or app is doing. What's your go-to for spotting logic patterns? #ReverseEngineering #Assembly

Stop wasting time on "variable_123" and "a[0x1]." From string-array decoders to control-flow flattening, learn how to untangle modern obfuscation and actually read the source. Babel examples included. #JavaScript #ReverseEngineering

https://trynoguard.com/learn/javascript-deobfuscation-techniques

Stop arguing with AI that tells you it can't answer your reverse-engineering or malware analysis questions. Learn why the refusals happen and how Aether handles them without the moralizing layers. #malware #reverseengineering

https://trynoguard.com/learn/why-chatgpt-refuses-reverse-engineering

Used Aether to deobfuscate a mess of parameterized JavaScript and map it back to its original function logic in minutes. Worked perfectly.

What's the hardest piece of code you've unfaded this week? #reverseengineering #malwaresamples

Did you know Weight Locking in Llama 3 and State-of-the-Art models isn't enough to stop "model stealing"? Speculative Decoding allows attackers to reconstruct trained parameters through high-volume API queries. Can't lock a weights file? Just query your way to it. #LLM #ReverseEngineering