Julián Arce 🇦🇷

🇦🇷 Coronados de gloria vivamos O juremos con gloria morir 🇦🇷

[H4RTO @SiempreHARTO]

Sospecha tranquilo porque hasta mañana yo no pienso hacer nada.

[Carlos Maslatón @CarlosMaslaton]

Para que quede claro, yo sí nacionalizo y le atribuyo todo sentido histórico y político a la semifinal de mañana Argentina-Inglaterra. Lo entiendo a Scaloni, él es profesional del fútbol y no puede hablar, pero yo sí puedo hacerlo porque soy un mero hincha. Para mi aquí se juega la contestación argentina a la expulsión de Rattín en Wembley en 1966 y el partido bombeado por el referee alemán. Se juegan las Invasiones Inglesas de 1806 y 1807. Se juega el crédito fraudulento de la Baring Brothers y la Vuelta de Obligado. La usurpación de las Malvinas y la memoria de nuestros heroes de 1982. No es solo fútbol, es el honor argentino, lo es todo, es la Patria misma. El apoyo a la Selección Argentina debe ser absoluto. Las críticas están completamente fuera de lugar en momentos nacionales como este. Ellos nos representan y saldrán a vencer o morir.

[La Sagrada Família @sagradafamilia]

Veiem per primera vegada la torre de Jesucrist il·luminada!
L'espectacle de llum iniciat des de la base fins a la il·luminació de la creu ha culminat amb una composició de llums guiats per drons, que han dibuixat la figura Gaudí i la frase «primer l'amor, després la tècnica».

[RJ 🟩 @rj_aligned]

first mobile demo of our waas:

same flow as before, built with our react native sdk.

- create an account and a passkey-backed smart wallet, signed with face id
- add a new passkey
- recover your account

[RJ 🟩 @rj_aligned]

new demo of our waas at @alignedlayer: from creating an app to recovering an account, all on passkey-backed smart wallets.

- sign up as a developer and get org + app + publishable key

- create a user wallet: a passkey on the device (touch id) backs a smart-contract wallet, with a server-held guardian as a second key

- send a transaction signed 2-of-2 (passkey + guardian)

- add a new passkey, approved on-chain by an existing one

- recover an account: the server adds a new passkey under a timelock.

[Julián Arce 🇦🇷 @JulianFArce]

@PabloSabbatella Como validan que seas vos realmente? Las veces que interactue con operadores con solo tener el DNI podia hacer cualquier cosa

[Facundo🇦🇷✈️ @facuc150]

Volvieron los Twin Otter a la Antártida, veremos si se trata de una estadía pasajera o se quedan permanentemente en la Base Marambio

[Aligned @alignedlayer]

ALIGN Token Sale KYC Registration: Extended One Week

Registration for the ALIGN Token Sale on Sonar is now open through 21st of April 7pm UTC.

We've heard from community members still completing verification. Rather than cut anyone off mid-process, we're extending the KYC window by one week.

The sale parameters remain unchanged:
- English auction using Sonar's technology
- 1% of total supply (100M ALIGN)
- Starting price $0.01
- Ceiling price: $0.1

sale website in the post below 👇

[Aligned @alignedlayer]

https://t.co/3LVVKr2fYf

[Aligned @alignedlayer]

https://t.co/tFaTOy3AJn

[P-OPS Team @POpsTeam1]

🚨 ZK Arcade Airdrop Update

@alignedlayer’s ZK Arcade allocations are now live in the checker.

🧩 Distribution split:
• NFT holders
• Leaderboard points

���� ZK Arcade activity snapshot:
• 5,644 proofs verified
• 522 NFTs minted
• 657.22M points claimed

👀 Check your allocation + full breakdown:

🔗 Tokenomics Update: https://t.co/uxLPFwjyUa

🔗 Airdrop Checker: https://t.co/nq0IiSEEWv

⚠️ Token is NOT live yet — this is informational only.

🔜 Next steps + claim instructions coming soon.

If you participated, now’s the time to verify your allocation 👇

[Federico Carrone @federicocarrone]

LLMs now make critical decisions in hospitals, defense, banks, and governments. Yet nobody can verify which model actually ran, or whether the output was tampered with. A provider or middleman can swap weights, silently requantize the model, alter decoding, inject hidden prompts, do supply chain attacks or change the deployment surface without the user knowing.

This problem is already serious. It will become critical.

We think this needs a practical solution, not just a theoretically clean one. CommitLLM is designed to be deployable on existing serving stacks now: the provider keeps the normal GPU serving path, does not need a proving circuit, does not need a kernel rewrite, and does not generate a heavy proof for every response.

In practice, two families of approaches dominated the conversation before this work: fingerprinting, which can be gamed, and proof-based systems, which are theoretically strong but too expensive for production inference.

We built CommitLLM to target the middle ground.

The core idea is to keep the verification discipline of proof systems, but specialize it to open weight LLM inference. The cryptographic core is simple: Freivalds style randomized checks for the large linear layers, plus Merkle commitments for the traced execution. Then a lot of engineering work is needed to make that line up with real GPU inference.

The key trick is this. A provider claims `z = W × x` for a massive weight matrix. Normally you would verify that by redoing the multiply. Instead, the verifier samples a secret random vector `r`, precomputes `v = rᵀ × W`, and later checks whether `v · x = rᵀ · z`. Two dot products instead of a full matrix multiply. In the current implementation, a wrong result passes with probability at most `1 / (2^32 - 5)` per check.

A full matrix multiply, audited with two dot products.

Most of the transformer can then be checked exactly or canonically from committed openings. Nonlinear operations such as activations and layer norms are canonically re executed by the CPU verifier. The one honest caveat is attention: native FP16/BF16 attention is not bit-reproducible across hardware. CommitLLM verifies the shell around attention exactly, then independently replays attention and checks that the committed post-attention output stays within a measured INT8 corridor. So attention is bounded and audited, not proved exactly.

That means the protocol already gives very strong exact guarantees on the parts that matter operationally most. If an audited response used the wrong model, the wrong quantization/configuration, or a tampered input/deployment surface, the audit catches that exactly. That includes things like model swaps, silent requantization, and provider-side prompt or system-prompt injection.

Today the implementation and measurements are strongest on Qwen and Llama. But the protocol itself is not meant to be Qwen-or-Llama-specific: we expect it to generalize across open-weight decoder-only families. What still has to be done is the engineering work to integrate and validate more families explicitly, and we are already working on that.

On the measured path, online generation overhead is about 12–14% with the provider staying on the normal GPU serving path. The heavier receipt finalization cost is separate and can be deferred off the user-facing path. The main systems costs are RAM and bandwidth, not proof generation.

The full response is always committed, but only a random fraction of responses are opened for audit. Individual audits are much larger, roughly 4 MB to 100 MB depending on audit depth. The important number is the amortized one: under a reasonable audit policy, the added bandwidth averages to roughly 300 KB per response.

After too many weeks without sleep, I’m proud to show what I built with @diego_aligned: CommitLLM.

The code and paper still need some cleaning and formalization. We're already in talks with multiple providers and teams that have cryptography related ideas on how to improve it even more. We're really excited about this and we will continue doubling down on building products in AI, cryptography and security with my company @class_lambda.

If governments, hospitals, defense and financial systems are going to run on LLMs, verifiable inference is not optional. It is infrastructure.

[Aligned @alignedlayer]

https://t.co/ALnr65DVAb

[Aligned @alignedlayer]

https://t.co/Jjur1BdfZi

[Aligned @alignedlayer]

NEW PRODUCT DEMO LIVE

Introducing Aligned's Wallet as a Service

[RJ 🟩 @rj_aligned]

wanted to share our first public demo of @alignedlayer wallet as a service.

our engineering team has been doing an incredible job. really proud of how fast we’re building this.

this is a key product for us. it’s how we help partners like @sovraio onboard millions of users in a seamless and secure way.

still work to do, but we’re on the right path. here’s the demo:

1. sign in with google

2. create a wallet with touch id passkey saved to icloud

3. send usdt on chain signed with your fingerprint

4. open on a new device

5. send another tx immediately

how we’re building it:

- 2 of 2 multisig where user passkey and server co sign every transaction

- erc 4337 account abstraction

- auth via google + webauthn passkeys.

- passkeys sync across devices via icloud keychain or google password manager

- server guardian recovery with timelock + mfa

still under development:

censorship resistance allowing users to retrieve their encrypted server key independently using their device key + mfa so no one can block access to their wallet

we will be sharing more demos in the future and a full blog post explaining the architecture and design decisions.

[RJ 🟩 @rj_aligned]

.@alignedlayer Proof Aggregation Service is now live on Ethereum Mainnet (alpha release).

from now on, our Rollup as a Service clients will
pay a lot less for L1 settlement.

demo below: watch a zk proof from @SuccinctLabs SP1 get aggregated through our system.

why this matters: in a future with hundreds or thousands of rollups settling on @ethereum, proof verification becomes a major cost. services that allow rollups to reduce their operational costs through aggregation are essential for that future to work.

current support:
- SP1 from @SuccinctLabs: live now
- @ethrex_client L2: integrated
- @ziskvm: integration in progress
- our zkVM: coming soon

links to verify the txs below.

[Aligned @alignedlayer]

https://t.co/Dwh2QBSSrC

[RJ 🟩 @rj_aligned]

demo of the new version of @alignedlayer zk proof aggregation service is now live on hoodi.

with this new version the proof aggregation service and the proof verification layer (avs) are two separate products.

users can now send proofs to one service or the other independently.

we will publish a blog post and update the docs with the new architecture for those interested in the details.

in the video you can see:

1. a user submitting an sp1 proof to the new gateway service.

2. the aggregation mode being spun up and looking for pending proofs.

3. multiple sp1 proofs being aggregated.

4. a single aggregated proof being submitted to @ethereum for verification.

this is the tx hash of the aggregated proof being verified. 0x32a8488bc4637cbacfadf51bcf9b3c13db9dd2582da09392e959d4864f14bf23.
you can check the etherscan and aligned explorer links in the post below.

great work by @mauro_aligned and team.

[Ethereum on ARM (and RISC-V) 🦇🔊🐼👉👈🐼 @EthereumOnARM]

Big news for Ethereum on Arm!

We've successfully completed the first sync of ethrex (@ethrex_client) on an Orange Pi 5 Plus, using the official ethereum on arm image.

Total sync time: 9 h, 16 m.

This is a fantastic milestone for running a minimal EL client on Arm.

1/3