M.A Dhavan retweeted
We've posted a 156-page tech report on Beam! We have good systems and it's worth reading. It covers all phases through pre-training, mid-training, SFT, RL, and OPD and goes into the infra, data, and scientific decisions we made throughout the process
📚 reflection.ai/documents/beam…
We've released the weights for Beam!!
🤗 huggingface.co/reflection
📄 Model card: reflection.ai/beam-model-car…
🧪 Try Beam in our playground & get an API key: platform.reflection.ai/playg…
📚 AND we take open model safety seriously and have posted our safety report: reflection.ai/beam-safety-re…
M.A Dhavan retweeted
Let's be precise about what the Epstein network actually was.
Not a "sex scandal." Not a tabloid story about a weird billionaire and his famous friends. Not a trafficking case in the sanitized, clinical language that makes it sound like a logistics problem.
It was a blackmail infrastructure for the ruling class.
Epstein didn't accumulate powerful men by accident. He accumulated them by design.
The private island. The private jet logs. The hidden cameras. The underage girls.
This was an intelligence operation.
A systematic, decades-long project to compromise the people who run governments, banks, media companies, universities, and courts.
And when he was arrested the second time, when it was finally impossible to make it disappear:
He died in federal custody.
Suicide, they said.
In a cell where the cameras malfunctioned.
With guards who had both fallen asleep.
After his previous suicide attempt had already put him on watch status.
The medical examiner hired by his brother found evidence consistent with homicide. The official medical examiner ruled it suicide.
And the name list, the full, unredacted list of everyone Epstein brought to that island and those houses and those parties, remains sealed.
Protected by the same judicial system that is supposed to protect you.
This is not a conspiracy theory.
The documents, the flight logs, the depositions, the testimony, all of it is real, much of it is public, and what it shows is a power structure that protects itself from accountability at every level.
The people on that list run your country.
They run your media.
They run your courts.
They are still running everything.
And they will call you a "conspiracy theorist" for noticing.
For anyone working on RSI via weight editing, here’s more rigorous prior work worth studying/citing:
1. Arditi et al. Refusal in Language Models Is Mediated by a Single Direction.
2. Meng et al. Locating and Editing Factual Associations in GPT.
3. Hase et al. Does Localization Inform Editing? Surprising Differences in Causality-Based Localization vs. Knowledge Editing in Language Models.
4. Geva et al. Transformer Feed-Forward Layers Are Key-Value Memories.
5. Ilharco et al. Editing Models with Task Arithmetic.
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math.
rikiyat.github.io/blog/weight-sp…
github.com/RikiyaT/lm_und…
huggingface.co/Qwen/Qwen3.5-0…
huggingface.co/RikiyaT/Qwen3.…
M.A Dhavan retweeted
This is an old hat. That's exactly what the neural fast weight programmers did in the 1990s (including the 1991 unnormalised linear Transformer). Overview with links to the original papers: people.idsia.ch/~juergen/fas…
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math.
rikiyat.github.io/blog/weight-sp…
github.com/RikiyaT/lm_und…
huggingface.co/Qwen/Qwen3.5-0…
huggingface.co/RikiyaT/Qwen3.…
M.A Dhavan retweeted
Elon Musk
Just what an incredible fall from grace…
From innovative, eco-conscious, cool nerd…
To depraved, entitled, evil, war-mongering, self-radicalised fuckwit.
M.A Dhavan retweeted
The University of Sankore in Timbuktu was operating in the 12th century.
At its peak it had 25,000 students.
It held between 400,000 and 700,000 manuscripts on mathematics, astronomy, medicine, law, literature and theology, many of which are still being catalogued today.
Seven hundred thousand manuscripts.
In a city that European maps labeled "terra incognita."
Unknown land.
Not unknown to the scholars studying astronomy inside it.
Not unknown to the merchants who traveled thousands of miles to reach its markets.
Not unknown to the students who came from across the continent and the Arab world to study there.
Unknown only to the people who needed it to be.
Because 700,000 manuscripts changes the sentence.
Because a university operating a thousand years ago in sub-Saharan Africa changes every assumption that colonialism was ever built on.
And the assumptions needed to survive.
So the manuscripts were not taught.
And the "terra incognita" remained, in the minds of those who needed it to, incognita.
M.A Dhavan retweeted
In this (extremely long) video, I'll build a distributed training framework from first principles using PyTorch.
We'll derive everything from first principles, explaining the mathematics behind distributed training as well as every auxiliary component we build along the way, including Multi-head Latent Attention (MLA), Rotary Position Embeddings (RoPE) and YaRN.
I'll explain process groups, ranks, collective communication operations, distributed autograd, device meshes and distributed tensors, and then use these foundations to implement the major forms of parallelism used to train modern language models.
We'll derive and code pipeline parallelism, data parallelism, FSDP, HSDP, tensor parallelism and context parallelism, including pipeline schedules, block matrix multiplication, model sharding and ring attention. We'll then combine these techniques and examine how tensor layouts and gradients must be tracked as data moves between devices.
Most importantly, we won't study each parallelism technique in isolation. We'll combine pipeline, data, tensor, context and expert parallelism into a single working framework and follow the tensors, communication operations and gradients as they move across GPUs. By the end, you'll understand not only how these techniques work individually, but how modern distributed training systems compose them to train large dense and mixture-of-experts models.
We'll also build the important components of a modern transformer, including rotary embeddings, long-context extensions, multi-head latent attention and a mixture-of-experts architecture. This leads into expert parallelism, expert tensor parallelism, token dispatching, all-to-all communication and combining different parallel dimensions through dense and sparse device meshes.
We'll also build the supporting training infrastructure, including the training loop, optimizer, learning-rate schedule, checkpointing and metrics.
The goal is to give you the foundations required to understand how distributed training frameworks work internally and how to design and implement new parallelization strategies yourself. No prior knowledge of distributed training is required.
Link to the video: youtube.com/watch?v=XoGvCBRn…
M.A Dhavan retweeted
🚀 Introducing DeepInfer, our in-house inference framework, now fully open source!
Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable.
🔗 github.com/deepseek-ai/DeepI…
1/4
M.A Dhavan retweeted
ICYMI an Israeli company hijacked online ads to track millions of devices and open the door to zero-click spyware, with our nuclear science agency, critical infrastructure and newsrooms compromised. This is a big deal.

ALT ABC news screenshot Australia's nuclear agency targeted with ads linked to spyware that infect devices without a click By national AI reporter Cam Wilson Topic:Cyber Security Posted about 14h agoabout 14 hours ago, updated about 13h agoabout 13 hours ago The ads appear to be a generic image that says "SALE" and shows a picture of miscellaneous products. The ads appear to be a generic image that says "SALE" and shows a picture of miscellaneous products from Ugreen, a Chinese consumer electronics company. (X: John Scott-Railton) In
M.A Dhavan retweeted
Australian police are rolling out Cellebrite's Israeli phone cracking software, and now another Israeli tech firm has turned online ads against our agencies. We need an urgent review of all surveillance companies operating here and the risks they pose to privacy and public data.
M.A Dhavan retweeted
We have released a free open source tool 'Zero Breach' - that can identify whether a recent Israeli spyware campaign left traces on your PC - the one Citizen Lab exposed, where Israeli ad-tech was used to infect journalists through 'no-click' ordinary ads.
It needs Python installed (python.org — tick "Add python.exe to PATH" during setup). Then open PowerShell in the folder where you saved it and run:
python -X utf8 zero-breach.py
It takes about 15 minutes. It only reads - it changes nothing, needs no admin, and never connects to the internet. It checks for 338 indicators and tells you if one of your browsers interacted with one of the infected ads in the past. (It also checks for Candiru infection)
Download it here: github.com/interplains/zero-…
You can also check whether your browser or device already blocks these kinds of ads, without downloading anything, here:
archivegenocide.com/scan/
You are looking for a green result ✅ - if not, fix your browser settings, install an ad blocker / DNS resolver.
For phones: Tens of thousands of apps show these ads, so we suggest you set up a DNS resolver (NextDNS, AdGuard, or a router/Pi-hole blocklist) so your phone blocks all of them automatically.
We also publish ready-made blocklists you can subscribe to as well. Use them.
Pease report any issues in our telegram linked in bio
God bless.
M.A Dhavan retweeted
All our problems basically come down to narrative control.
People are manipulated into believing the story that Israel's existence must be preserved at all cost because of this religious belief or that political ideology.
People are manipulated into believing the story that western elections are real and that voting in the two-party system is the only path toward beneficial policies.
People are manipulated into believing the story that capitalism is the best possible system, and that socialism always fails (organically, all by itself) because its tenets go against human nature.
People are manipulated into believing the story that the US-led liberal world order, while perhaps flawed, is still our best shot at a just and stable world.
People are manipulated into believing the story that the mass media are telling us the truth about our society and the world, and that everything works pretty much the way we were taught in school.
People are successfully manipulated in these ways because the wealthiest among us control the media and own all the largest platforms, which gives them a very high degree of narrative control. They decide what stories get told to the largest audiences about how the world works, so they control public perception of reality.
This has been happening for generations, so we grow up with our parents, teachers, preachers and pundits teaching us the narratives that they were trained to believe when they were young.
So we grow up marinating in this. It's so rife throughout our society that most people don't even notice it's happening; they just think all those narratives are what reality objectively looks like.
But they're not. They're just thoughts. Mental noises. Stories made of words. Pick any of them apart and you'll find that's all they are: empty narrative fluff. Their power comes entirely from large numbers of people imbuing those narratives with the power of belief.
The consciousness of the average human is dominated by mental stories, so if you can control the mental stories the average humans are telling themselves, then you can control the humans. Get them buying into your preferred narratives about reality and you can exert huge degrees of influence over the way they think, speak, work, shop, live, and vote.
Our entire civilization looks the way it looks because of this vast network of influence. It's what made the Gaza holocaust possible, what makes homelessness possible, what makes ecocide possible, what makes war and militarism possible. If the public were not being successfully manipulated into accepting the narratives which justify these realities, then everyone would start un-jacking their minds from the psychological control network and building the kind of world they would prefer to see.
And what's interesting is that believed mental narrative is also the source of suffering in humans as individuals. The only reason you're unable to be present in the moment and rest in peaceful contentment with a serene mind is because of all the thought-stories about reality that you are holding onto inside yourself, both consciously and unconsciously.
As we grow up and experience traumas in early childhood, we start forming beliefs about reality and who we are and how we need to be, which we then carry in our bodies deep into adulthood. With a lot of inner work we can begin releasing these old beliefs and freeing up the energy our bodies have been feeding into them this whole time, liberating ourselves from the internal commotion and dysfunction they cause.
"I am unworthy."
"I am deficient."
"The world is a hostile place and I need to protect myself from it."
The young human organism forms all these belief constructs about itself and its reality, and then those beliefs quickly become maladaptive because they were built in youthful naivety without taking real-time feedback from life's ever-changing circumstances. It is only by becoming aware of them and consciously removing the power of belief from them that they release their grip on our lives.
It's a challenging process, but if you take it deep enough it will completely transform your entire experience of reality in some surprising and delightful ways. You will discover that even your most fundamental assumptions about what you are, what your world is, and how your experience is happening were never anything more than believed thoughts held on a subtle energetic level, and that life is so much simpler and more peaceful than you had ever imagined it could be.
Our collective untangling from believed mental narrative is unfolding in the same way. Gradually we are learning together that everything we were indoctrinated into believing about our world and how it's happening was completely false from top to bottom, and we are beginning to deconstruct our power-serving worldviews one by one.
The more this happens, the weaker the hold our rulers will have on our minds at mass scale. The less we buy into the old narratives, the less capable the manipulators will be of influencing our thoughts, words and actions to their advantage.
And from there, anything is possible. If human consciousness becomes liberated from believed mental narrative on all levels, we could truly have paradise on earth.
I am more convinced then ever that this is possible. I am more convinced than ever that it is already happening.
M.A Dhavan retweeted
Veteran journalist Mujahid Abu Mufleh, before and after his detention in the Israeli Auschwitz camps.
M.A Dhavan retweeted
It's weird to me that Aussie journos like myself, Antoinette, Antoun Issa, Joel Jenkins, Jan Fran, Caitlin Johnstone, to name a few, have been proven correct on Israel/imperialism over and over, & still ppl listen to the same Murdoch/ABC hacks that have been lying this whole time
M.A Dhavan retweeted
People seem confused about why Tesla is getting so much backlash over its new "flying Roadster." For those of us who were once genuinely enthusiastic about Tesla, this isn't about hating innovation. It's about a long history of broken promises, misleading expectations, and increasingly absurd marketing.
Tesla unveiled the second-generation Roadster in 2017, took customer deposits, and promised deliveries in 2020. Six years past that deadline, it still hasn't delivered the car. Instead, it's demonstrating an even more outrageous version with features that will likely never make it into production. The original 2nd gen Roadster would still be an amazing car today. Why not just deliver what was promised?
Then there's Full Self-Driving. Tesla marketed the idea that Model 3 owners could turn their cars into income-generating robotaxis. That never materialized. Even if the technology eventually gets there, the opportunity has already passed for many early buyers. Timing matters. Delivering something a decade late isn't fulfilling the promise you sold to customers.
And then there's Solar Roof. Unveiled in 2016 as a revolutionary approach to residential solar, it struggled with costs, installation challenges, and adoption far below expectations. After nearly a decade, Tesla discontinued it entirely in August 2026. Another grand vision abandoned, leaving existing customers with a discontinued product to maintain.
Then came the Cybertruck, marketed as a nearly indestructible, revolutionary truck. Instead, we got a vehicle with questionable build quality, numerous recalls, expensive repairs, disappointing results in several real-world off-road tests, and specifications that fell short of the original promises. It literally looks like a dumpster, but aesthetics aside, the collapsing demand speaks volumes about the reality of what Tesla delivered. Another heavily hyped product that fell dramatically short of expectations.
This is how consumer trust erodes. Not through one disappointment, but through years of accumulating them. And this is part of the reason why many people who were once enthusiastic about Tesla have become critical and cynical about it.
But there's something even more fundamental here: Tesla has completely departed from the grounded, practical vision that made so many of us enthusiastic about the company in the first place.
Electric cars were ambitious, but the idea was grounded in reality. People could charge at home, infrastructure could gradually expand, technology could improve, and millions of ordinary people could eventually benefit. There was a realistic path toward mass adoption.
What the hell does a flying Roadster powered by compressed-air thrusters have to do with any of that?
It's an expensive, impractical engineering stunt with enormous packaging, safety, and regulatory challenges, and no credible near-term path toward mass-market usefulness. Modern electric cars already achieve ridiculous acceleration without this nonsense. What practical problem does it solve? Who actually needs it? When will ordinary consumers benefit?
And watching Tesla fans compare this experimental prototype in a Roadster-shaped shell to actual production vehicles people can buy today is ridiculous.
What's especially infuriating is the cult-like mentality surrounding all of this. You can't point out the history of failed promises, question the practicality, or demand realistic timelines without being accused of being a "hater" or having "Elon derangement syndrome."
Tesla deserves recognition for helping make electric cars a reality, but that success doesn't give it a blank check to keep marketing increasingly fantastical promises while failing to deliver on previous commitments.
This isn't hatred. It's disappointment from people who once believed in the company, watched it accomplish extraordinary things, and have since watched it increasingly substitute practical innovation with spectacle and promises.
Elon fanboys are also ignoring a bigger pattern that goes beyond Tesla.
What made many of us enthusiastic about Elon and Tesla and in the first place was a grounded vision of technological progress that could genuinely improve millions of lives. Electric cars were ambitious, but practical, achievable, and useful to ordinary people. Now we're supposed to get excited about flying cars with compressed-air thrusters that have no credible path to mass adoption and solve no meaningful problem for the average consumer.
It's the same kind of thinking behind marketing the colonization of Mars as some urgent necessity for humanity's survival. There's nothing wrong with space exploration, but Earth isn't about to become uninhabitable because of the Sun. We're talking about geological timescales, while humanity faces enormous problems RIGHT NOW: poverty, hunger, access to education, corruption, abuse of power, wars, and a lack of government accountability. We have plenty of very real problems to solve before worrying about humanity's survival on another planet.
And then there's DOGE. I was genuinely excited about the original idea of making government more efficient, reducing waste, and increasing accountability. Instead, we got questionable savings figures, inexperienced people making sweeping decisions, serious concerns over access to sensitive personal information, and an initiative that ultimately fizzled out without delivering its grand promises. Another enormous narrative that fell apart when confronted with reality.
That's what makes all of this so frustrating. It's not that we stopped believing in technological progress. It's that we've watched someone who once pursued grounded, transformative ideas increasingly embrace fantastical promises, political spectacle, and questionable narratives disconnected from practical reality.
And somehow, pointing out this history of disappointment gets you labeled a "hater" or accused of having "Elon derangement syndrome."
The problem isn't dreaming big. It's repeatedly selling people those dreams, failing to deliver, and then expecting everyone to applaud the next, even more ridiculous promise without asking what happened to the last one.
The world doesn't need more spectacular promises. It needs practical solutions to real problems, honest expectations, and accountability for what was promised. That's what made Tesla exciting in the first place, and that's what so many of us feel has been betrayed.
M.A Dhavan retweeted
It is not enough to recognize Palestine.
Because recognition still assumes "Israel" has a right to exist as it is.
A colonial regime cannot coexist with the people it was built to erase.
So long as "Israel" is recognized, Palestine will never be free.
True justice begins not with recognition, but with un-recognition of the criminal project that calls itself a state.
M.A Dhavan retweeted
Finally, clear, unequivocal, irrefutable, and credible proof that Israel shoots children in Gaza. Impossible to deny.
Pure evil.
M.A Dhavan retweeted
Maryam Ahmed Mahmoud Ghanem lies in a hospital bed with severe burns, clutching her doll — the sole survivor after a strike killed her entire family, including her mother.
M.A Dhavan retweeted
🚨HORRIFIC UPDATE: The Palestinian mother Israel reportedly burned alive in Gaza has been identified as Reem Awad.
Her little daughters reportedly WATCHED THEIR MOTHER BURN TO DEATH. ONE OF THEM HAD ALREADY LOST HER LEG IN A PREVIOUS ISRAELI BOMBING.
ISRAEL TOOK HER LEG. NOW ISRAEL HAS TAKEN HER MOTHER.
MY GOD. WHAT MORE MUST THESE CHILDREN SUFFER AT ISRAEL’S HANDS?!



























