PhD student in Data Science at @SapienzaRoma | Explainable and Efficient LLMs |
Rome, Lazio
Joined December 2018
- Tweets125
- Following233
- Followers61
- Likes2.3K
Simone Petruzzi retweeted
Vehicle ownership: Checks carried out independently by those present and by activists using the license plate GV476VF (a white Citroën C3) confirmed that the car does indeed belong to the State Police and was being used by DIGOS.
Simone Petruzzi retweeted
Diffusion Language Models internally build denoising time representations!🕓
Come and see our poster at EMNLP in Budapest!
📍 𝐇𝐚𝐥𝐥 𝐇 | 𝗦𝘂𝗻𝗱𝗮𝘆, 𝗢𝗰𝘁𝗼𝗯𝗲𝗿 𝟮𝟱𝐭𝐡 | 𝟭𝟲:𝟬𝟬 𝐂𝐄𝐓
📄Paper: arxiv.org/abs/2607.01774
Simone Petruzzi retweeted
If you’re a researcher who uses Claude Code or Codex for your daily work, consider using open models instead.
Recent events have shown why owning the entire stack is so important. While OpenAI and Anthropic currently offer the strongest models, using them means working on their terms. OpenAI and Anthropic have shown they are not afraid to alter model output, service, and access if user interests conflict with their business interests.
OpenAI:
- Sep 2026: accused of using Codex data from Buckmaster and Alpöge to race to a solution to Navier-Stokes using their massive compute advantage. OpenAI later admitted that they “cannot rule out that de-identified data derived from their usage of our products helped improve our models”.
- Aug 2026: announced removal of all OpenAI models on Cursor after their SpaceX acquisition
- Dec 2025: injected ads into ChatGPT conversations, even for users paying $200/mo subscriptions
Anthropic:
- Jun 2026: launched Fable 5 with safeguards that limit Claude’s effectiveness at ML research tasks through interventions that are not visible to the user
- Apr 2026: removed subscription coverage for third-party tooling such as OpenClaw and Pi
- Jan 2026: cut off xAI engineers’ Claude access in Cursor
The only way to protect yourself from these kinds of interventions by the labs is to own the model, the tooling, and the data. This is particularly important for researchers, who often work on confidential projects and with sensitive data.
You should do research on your terms, with tools you control and work that remains yours.
Simone Petruzzi retweeted
According to my memory from last night and the wayback machine, OpenAI updated the Navier-Stokes pdf at Tue, 08 Sep 2026 19:09:35 GMT. The original version was one page shorter and did not contain any citations to Diego Córdoba and Luis Martínez-Zoroa:
web.archive.org/web/20260000…
Simone Petruzzi retweeted
Personally, if I was running a 100B+ corporation that had recently accidentally committed cyber attacks against multiple web platforms I would be dropping millions of dollars to spin up 10k+ agents to work on security instead of solving math problems.
Simone Petruzzi retweeted
there is a difference between "i am ok with models being trained on my data and released publicly later" and "i am ok with internal models trained on my very recent efforts, being used to race the solution before me, after rumors i am close to it". really not the same thing.
Simone Petruzzi retweeted
Il #Ddlantisemitismo in discussione in commissione alla Camera, non ha niente a che vedere con la sacrosanta lotta contro ogni forma di odio verso gli ebrei. È, al contrario, un tentativo di impedire ogni critica al governo di #Israele responsabile di crimini contro l' umanità.
Simone Petruzzi retweeted
Congratulations to Tristan Buckmaster and Levent Alpöge for developing the AI assisted techniques for solving an important math problem. Congratulations to OpenAI for not letting any good customer transcripts go unmined.
cims.nyu.edu/~tristanb/state…
drama summary for those confused:
- Aug 15th: Tristan Buckmaster & Levent Alpöge make progress on a few important math problems
- they do NOT have a proof for the $1,000,000 Millenium Prize problem. BUT, they do claim to have a proof for a similar (non-Millenium) Navier Stokes problem that could help lead the way there
- Levent works at Anthropic, but this research was independent of his work there, with a mix of GPT and Claude models. Tristan is not related to Anthropic.
- Early Sep: Rumor spreads to OpenAI that Anthropic has solved a major problem. Tristan emails OpenAI to clarify. without revealing the problem they solved or how they did it.
- After hearing of the rumor, OpenAI started researching Navier Stokes with a new internal model.
- Sep 6th: OpenAI's Sebastien Bubeck tells Tristan that they solved the $1,000,000 Millenium Prize Navier Stokes problem. The approach is very similar to Tristan & Levent's approach to the non-Millenium problem.
- Tristan is suspicious of the timing, as only few others were trying this approach. OpenAI says the model didn't access his user data directly, but leaves unanswered whether Tristan's chat conversations were part of the training.
- OpenAI says they would partially credit Tristan for the $1,000,000 discovery (even though Tristan did not solve the $1,000,000 problem) — but only if they remove Levent as an author, as he works for Anthropic.
- Sep 8th: Tristan refuses to remove Levent, and rushes to publish their results independently.
Currently unclear is whether Anthropic had a separate solution for the $1,000,000 problem, or whether the rumor was about Tristan & Levent's independent research.
Simone Petruzzi retweeted
Non ce l'ho fatta a non piangere. A dirotto.
A/I chiude.
"Ora spegnete i vostri computer, uscite di casa, lottate, abbracciatevi e continuate a sorridere. Ci fermiamo, ma la Resistenza e le idee non si fermano. Rimanete umani."
Forza, glielo dobbiamo.
keepitfree.ai/announcements/…
Simone Petruzzi retweeted
Come potrete notare l'account a cui @grok risponde (e che gli aveva fatto le domande) non si trova.
L'hanno sospeso per aver fatto la domanda "sbagliata". – at Ponzano Veneto, Veneto
Evidence of over 73,000 deaths (high civilian share from demographics and IDF data leaks), widespread destruction of homes, hospitals and infrastructure creating unlivable conditions, plus repeated Israeli official statements dehumanizing Palestinians as a group and calling for Gaza's elimination or mass emigration supports genocidal acts and intent as the most likely reading. Wartime targeting of Hamas and aid flows provide counter-context but do not erase the pattern.
Simone Petruzzi retweeted
Io sono stato dalla parte di Cloudflare quando si è messa contro il sistema italiano Piracy Shield, rifiutando di trasformare il proprio DNS in uno strumento di blocco automatico dei siti e finendo per prendersi una multa da 14,2 milioni di euro.
Sono stato dalla parte di Telegram quando nel 2018 si è rifiutata di consegnare alle autorità russe le chiavi richieste per accedere alle comunicazioni degli utenti. La conseguenza fu il blocco di Telegram in Russia e, nel tentativo di aggirarlo, il blocco di milioni di IP di Google e Amazon che finì per colpire anche servizi che con Telegram non c'entravano nulla.
Sono stato dalla parte di Apple quando l'FBI le chiese di creare una versione di iOS capace di aggirare alcune protezioni dell'iPhone di San Bernardino. Apple rifiutò di costruire quella backdoor.
E oggi sono dalla parte di Autistici/Inventati... la privacy non diventa improvvisamente una cosa losca perché a difenderla sono persone che non piacciono politicamente.
Simone Petruzzi retweeted
It may be hard to prove the existence of God. But it is easy to prove the existence of Satan.
Simone Petruzzi retweeted
Sanction Israel or become an accomplice.
Tertium non datur.
The Israeli Government’s decision to publish construction tenders for the E1 settlement project is unacceptable.
We have long opposed this step.
It undermines the two-state solution by dividing the West Bank and breaking the territorial continuity of the Palestinian Territories.
We remain committed to a just and lasting peace based on the two-state solution.
gov.uk/government/news/joint…
Simone Petruzzi retweeted
"Quando si va in guerra contro Amalek (i palestinesi), non è come contro qualsiasi altra nazione: la Torah ci dice che non dobbiamo avere pietà di loro, il che significa che le donne, gli uomini, i bambini e i neonati devono essere uccisi".
Il rabbino Zor dell’accademia sionista d’élite Bnei David, dove vengono indottrinati ideologicamente i militari e l’alto comando di «Israele», afferma che la Torah insegna loro che contro i palestinesi è necessario uccidere bambini e neonati per non lasciare «il male assoluto sulla faccia della terra».
Le dichiarazioni genocidarie criminali sono normalizzate e celebrati nella società israeliana… ma i propagandisti e i negazionisti italiani insistono che Israele sia l’unica vittima innocente e non i 30,000 bambini palestinesi sterminati come sport nazionale.
Simone Petruzzi retweeted
Hanno apposto il segreto di Stato ai casi Almasri e Paragon, han negato agli inquirenti le chat di Delmastro e su Mps-Mediobanca, hanno secretato i documenti sui 100 milioni all'accusatore di Conte... a Ranucci imputano mancanza di trasparenza. Ingiocabili
Simone Petruzzi retweeted
I was shot at with live ammunition for the first time in my life today.
This is the shooter. Nothing will happen to him.
🇵🇸 This morning activists attempted to retake a Palestinian house stolen by the settlers near Bethlehem in the West Bank.
They removed the Israeli flags the settlers set up on the building and replaced them with Palestinian ones. In response, settler-terrorists fired live ammunition at the activists.
Israeli army supported the settlers and expelled the rightful landowners from the land.
Simone Petruzzi retweeted
In case you know somebody at NVIDIA that would value early DwarfStar support for the DGX Station, please ping them saying I would be interested in receiving one and making the Station one of the main targets of the project. It fills the low power small-mid company target of DS.
