awnyrvan

The community asked for it. Make Gains Great Again.

Pair listing ✨ @megaeth — The first real-time blockchain, where a new class of applications lives. Trade $MEGA with $USDM on @megaeth at: gains.trade/trading#MEGA-USD… Up to 150x leverage.

🚨BREAKING: Drift Protocol just got drained for over $200 million Solana's largest perps DEX. Gone in one transaction batch. The attacker didn't find a smart contract bug. They didn't exploit a flash loan. They walked in with the keys. On-chain data shows a single account initiating massive outbound transfers. SOL, JitoSOL, WETH, wrapped BTC, stablecoins in USD, EUR, and JPY. Even FARTCOIN. They took the FARTCOIN. A blockchain security researcher confirmed what everyone suspected: a private key compromise. The admin signer was either leaked or someone with access pulled the trigger themselves. And here's what makes it worse. The attacker funded the wallets a week before the exploit. Ran a test transaction. Then waited. This wasn't a hack. This was a heist with a rehearsal. Phantom Wallet already cut off access to the protocol. Drift posted about "unusual activity" and told users to stop depositing. $200 million gone and the official response is "unusual activity." Some estimates put the real number closer to $270 million. We won't know until the dust settles and the wallets stop moving. The funds are already being swapped to USDC and bridged to Ethereum. Classic exit playbook. This is potentially the largest Web3 exploit in three years. But who needs security when you have speed, right Kyle? 🚀

@reynardoew @awnyrvan @thedefiedge @Uniswap @pendle_fi @aave @HyperliquidX @SkyEcosystem @LidoFinance @sparkfinance @gains_trade @patfscott @0xngmi

Introducing 𝑨𝒕𝒕𝒆𝒏𝒕𝒊𝒐𝒏 𝑹𝒆𝒔𝒊𝒅𝒖𝒂𝒍𝒔: Rethinking depth-wise aggregation. Residual connections have long relied on fixed, uniform accumulation. Inspired by the duality of time and depth, we introduce Attention Residuals, replacing standard depth-wise recurrence with learned, input-dependent attention over preceding layers. 🔹 Enables networks to selectively retrieve past representations, naturally mitigating dilution and hidden-state growth. 🔹 Introduces Block AttnRes, partitioning layers into compressed blocks to make cross-layer attention practical at scale. 🔹 Serves as an efficient drop-in replacement, demonstrating a 1.25x compute advantage with negligible (<2%) inference latency overhead. 🔹 Validated on the Kimi Linear architecture (48B total, 3B activated parameters), delivering consistent downstream performance gains. 🔗Full report: github.com/MoonshotAI/Attent…

what stands out with $AVNT is: recent 4% airdrop supply was absorbed, the protocol is still generating real fees, and buybacks are now live. at around $48M market cap and $15M annualized fees, this looks like a solid value project.

Burn baby, burn. Recent 4% airdrop supply seems to have been well absorbed by the market. Meanwhile, @Avantisfi is still generating around $15M in annualized fees and $AVNT is holding up fairly well in the circumstances. Now add the fact that systematic buybacks are LIVE, with 30% protocol revenue directly going toward buying back and burning $AVNT At around $48M market cap, this looks like one of the more interesting R/R setups among Perp DEX and Base tokens right now.

We’re shutting down Polynomial Chain and Polynomial Trade. This one hurts to write. $4B+ in volume. 27M+ transactions. 22K of you believed in us. 8M TVL at peak. Here’s what happened, what we learned, and why we’re not done. 🧵