Mridul Arya

Building Full-Stack Engineer and Web(3) developer. Ex-Polkassembly.

[Mridul Arya @alpha_infinitus]

Here's what I've been building for the last few months.

It's called Blockpedia. You ask a question about Polkadot and it pulls answers from different ecosystem sources with citations so you can verify everything.

Still rough around the edges but it works. Would love honest feedback from anyone who deals with this data day to day or people new to the ecosystem.

https://t.co/89nK4EeFRr

[ClaudeDevs @ClaudeDevs]

https://t.co/sRyO7OigQ7 is our new home for developers building with Claude.

You'll find engineering deep dives, Claude Code and API guides, tips from the teams building Claude, and some fun easter eggs.

[BridgeMind @bridgemindai]

A lot of people are saying Anthropic has already nerfed Claude Opus 5.5.

We're launching NerfBench on BridgeBench tomorrow.

We have the day 1 results.

Tomorrow morning we show you the retest.

Is Claude Opus 5.5 nerfed or not?

[Krishna Y @joi_n9911]

Ten hypotheses go in. One comes out.
The other nine are paid for in full.

Sagitra is the research loop, automated. The nine bad ideas die in days instead of months.

Introducing your autonomous companion scientist.

100+ researchers at IIT Bombay, Imperial, Northwestern and Yale are already on it.

https://t.co/1XpAOhWQGV

[Rik Arends @rikarends]

51 million lines of code. Chromium codebase. Here you go. 120hz, 3gb ram use.

[Alex Cui @alexcdot]

Claude's watermark probably doesn't work how you think. As the CTO of GPTZero, I'll explain how Anthropic, Google and OpenAI are building text watermarking in this brief explainer and whether it can be defeated.

Almost all forms of watermarking that are fast and cheap enough for a frontier lab have the same formula, following the KGW method:

In generation:

1. Let's say you've generated n tokens so far. Take those n tokens + a secret key to generate a random hash
2. Use that hash to randomly reweight the probabilities for the n+1 token, and then sample from that new distribution. In the simple case, you could split 50% of all English words into a green or red set based on your hash, and boost the probability of words in the green set.

For watermark detection:

1. For each token, see if it was in the green or red set.
2. To do this, recreate the hash based on the secret key and the text preceding the current token. Then, recreate the green and red set of words.
3. Once you've checked all the words in the text, if the next token is selected disproportionally from the green set more than 50% of the time, you claim the text has the watermark.

I can tell you want to ask the following:

1) Isn't it easy to mess up the hash if you paraphrase the text? The answer is mostly yes, however, you can use a statistical model to get your hash instead of a deterministic function (SIR, Adaptive Watermark). Since the entire watermark is probabilistic, this is fine.

2) Doesn't this make the text much worse? The answer is yes, it does - Yes, it does – but for most people, it's imperceptible (Google claims in human feedback study with 20,000 texts), since there are exponentially many ways to write the same paragraph. DiPmark does something more sophisticated to avoid shifting the text distribution on average. Of course, watermarks fail on short text or highly predictable texts like "2+2=4".

3) Shouldn't it be easy to figure out the green and red sets? The answer is no. You would need an exponentially large number of samples from the watermarker to reconstruct those sets exactly, but it's a risk if the detector is open to the wild (Watermark Stealing)

Still, there are couple challenges that a frontier lab needs to overcome:
1. Their watermark needs to work token-by-token because they are streaming their text to users. Many watermark methods plan sentences or paragraphs at a time, or change the text after its entirely written, in order to make their watermark robust to paraphrasers, and a frontier lab cannot afford to do this yet (SemStamp, PostMark)
2. If the secret key leaks, the watermark is busted. To avoid a large blast damage from this, you need to have a couple secret keys in rotation.
3. There are some texts, like code, that cannot be arbitrarily changed, otherwise the code will break. In those cases, the watermark needs to selectively change words in parts of the text that can tolerate synonyms (i.e. like variable naming) - see SWEET, EWD, Invisible Entropy.
4. They will need to educate their users on how to deal with false positives and false negatives of a detector, which is a big challenge (one we put a lot of effort into)

So, how do I see this playing out in the next 6 months?
1. If Anthropic releases the watermark detector publically, I think they defeat their own watermark. People find reliable watermark removal strategies by testing against Anthropic (AI detectors like GPTZero have an advantage here because they can train against these adversaries once they become popular).
2. If they keep the detector private to the government, like Google has done, it's "safer". However, there are some papers showing trained approaches that work robustly to zero-shot break watermarks without any data, simply because they try to write the text just like a human (Zhang et al. 2024, Watermarks in the Sand). Also, making your detector makes it battle-tested and stronger long-term (my experience).
3. In my testing, the watermarks don't survive intense paraphrasing (especially if you combine word choice and syntax attacks), or human text substitution (rewrite your AI text by plagiarizing human authors). The free paraphrasers I've tried have quickly bypassed Google Deepmind's SynthId for what it's worth.
4. All-in-all, frontier labs are likely okay with this because they expect most users to not attack the watermark, and also because they + European regulators likely don't care past a certain point - its good enough.
5. Overall, I think users of frontier LLMs will not really care about this, because 1) they don't realize watermarks are there, 2) EU will force everyone to conform, 3) this seems more like regulatory hoop-jumping than an earnest effort from frontier labs to expose LLM use

Lastly, people's first concern shouldn't be watermarking, it should be AI detectors!

If you're posting, "its not X, its Y!!", I don't think the watermark is going to make a difference :)

[Krishna Y @joi_n9911]

Your research lives in 10 tabs that don't talk.
We built one place from question to submission. No fake citations. Ever.
https://t.co/YrtrmMMUx5

[ClaudeDevs @ClaudeDevs]

For the developers building with Claude, a direct line from the team.

Follow for changelogs, API releases, community updates, and deep dives.

[Cheng Lou @_chenglou]

My dear front-end developers (and anyone who’s interested in the future of interfaces):

I have crawled through depths of hell to bring you, for the foreseeable years, one of the more important foundational pieces of UI engineering (if not in implementation then certainly at least in concept):
Fast, accurate and comprehensive userland text measurement algorithm in pure TypeScript, usable for laying out entire web pages without CSS, bypassing DOM measurements and reflow

[Mridul Arya @alpha_infinitus]

Talk to the polkadot forum :
https://t.co/xs1QCDUBiY

[Mridul Arya @alpha_infinitus]

Wrote up more detail on how it works, why it exists, what data sources it covers on the Polkadot Forum:
https://t.co/OWthhbu7au

[Mridul Arya @alpha_infinitus]

Here's what I've been building for the last few months.

It's called Blockpedia. You ask a question about Polkadot and it pulls answers from different ecosystem sources with citations so you can verify everything.

Still rough around the edges but it works. Would love honest feedback from anyone who deals with this data day to day or people new to the ecosystem.

https://t.co/89nK4EeFRr

[kartikey singh @askwhykartik]

Here is a list of all startup accelerators you can apply to right now:

@ycombinator ($500k for ~7%)
@a16z ($750k-$1M for ~7-10%)
@pioneerdotapp ($20k for 1%)
@the_mint_vc ($500k for 10%)
@angelpad ($120k for 7%)
@techstars ($220k for ~5-7%)
@500GlobalVC ($112.5k for 6%)
@EFStartups ($250k for ~9%)
@southparkcommon ($400k for 7% + $600k guaranteed follow-on)
@sequoia ($1M)
@pearvc ($250k-$2M)
@greylock (SAFE note + $500k+ in credits)
@conviction ($150k uncapped MFN SAFE)
@openai ($1M equity investment)
@StartupWiseGuys (up to €65k for equity)
@apxaccel (up to €500k, typically €50k for 5%)
@southparkcommon ($150k for 5-10%)
@seedcamp (€100k-€200k for 7-7.5%)
@antlervc (€100k for 10% + stipend / $200k-$250k for 8-9%)
@googlestartups (up to $100k)
@accel (up to $500k-$1M)
@aigrant ($250k uncapped)
@aforecapital ($100k-$500k)
@BoostVC (up to $500k for 15%)

[Mridul Arya @alpha_infinitus]

@Syntropy Try it live in your browser:
https://t.co/ne0qz1oMOD
npm i @syntropy-labs/react-web-speech
⭐ https://t.co/4JyolnHnHm
#ReactJS #WebDev #BuildInPublic #TypeScript

[Mridul Arya @alpha_infinitus]

Was building an AI chat interface (like everyone is lol) and wanted voice input like the big boy apps..
didn't wanna pay for the speech to text for initial versions, and the Web Speech API is FREE in browsers but.. every package I found are just thin wrappers.
So I built it. 🎙️🧵

[Mridul Arya @alpha_infinitus]

@Syntropy One hook, zero config:
import { useSpeechInput } from '@syntropy-labs/react-web-speech'
const { toggle, transcript } = useSpeechInput()
That's it. Click to speak, auto-stops on silence.

[Polkadot @Polkadot]

we're sorry

[Nikhil Ranjan @niklabh]

Wrote a book on Ink! programming language @ink_lang . A memory safe way to write smart contracts.
https://t.co/lTvUlWPpE7

[Polkadot @Polkadot]

DOT gets you access to the world's largest DAO.

@polk_gov makes it easy to govern.

Polkassembly is a platform for engaging in OpenGov, bringing together all things on-chain in one place.

How to connect 👇