Telegram RegisterThe public register of Telegram
Telegram profile photo for • Dmitry Legchikov

Channel

• Dmitry Legchikov

@legchikov_ai

On this record: Growth · Engagement · Reactions · Posts · Citations · Cite this entry

820subscribers

+6 since we began measuring on 7 August 2026

Risers and fallers across the register · movement among entries of Under 1,000.

Register entry

Telegram ID-1001985961485
TypeChannel
Username@legchikov_ai
CreatedBetween 1 April 2023 and 31 October 2023 — estimated from Telegram’s id allocation, not measured. How this range is calculated.
First recorded8 August 2026
Last confirmed live29 August 2026
Measurements held5
Confirmed unchanged1 time, most recently 29 August 2026
On Telegramt.me/legchikov_ai

Growth

8108208157 August 2026 — 814 subscribers8 August 2026 — 814 subscribers8 August 2026 — 813 subscribers23 August 2026 — 810 subscribers29 August 2026 — 820 subscribers7 August 202629 August 2026
5 measurements spanning 22 days, net +6. Dots are measurements; the straight line between them is drawn to join them, not to claim we know the path taken in between — snapshots are recorded only when a count changes, so gaps mean “no change observed”, never “interpolated”. The vertical axis spans 809–822 and does not start at zero.
Measurement log — every subscribers count we have recorded
Measured (UTC)SubscribersChange
29 Aug 2026, 22:04820+10
23 Aug 2026, 20:24810-3
8 Aug 2026, 08:13813-1
8 Aug 2026, 01:33814no change
7 Aug 2026, 11:18814first reading

Engagement

18 posts held, back to 7 July 2026the reader has not yet reached the start of this channel’s public history, so older posts may sit further back, unread. Read across 1 page of Telegram’s post history, 20 posts per page.

ERR · 30 days
17.3%
avg views ÷ 820 subscribers
Avg views / post
142
6 posts measured
Reaction rate
3.05%
reactions ÷ views · ER floor
Posts in window
6
of 18 held

ERR is average views per post over the last 30 days divided by subscribers, the definition TGStat uses, so this figure is comparable with the one you will see elsewhere. It falls structurally as a channel grows: a high ERR on a small channel and a low one on a large channel describe reach mathematics, not quality. We publish the figure and the sample it came from and pass no verdict on it.

ER is defined industry-wide as (forwards + reactions + comments) ÷ views — note the denominator is views, not subscribers. Telegram’s public web preview carries views and reactions but not forward or comment counts, so the reaction rate above is the reactions term only and is therefore a floor: the true ER for this channel is higher by an amount we have not measured and will not estimate.

What these figures were computed from
WindowRolling 30 days · latest post in window 7 August 2026
Posts held18 (7 July 20267 August 2026)
Views total853
Reactions total26
Forwards / commentsnot exposed by the public surface — not measured, not estimated
Readings taken8 Aug 2026, 01:33 UTC

Views are a single reading per post, taken at the time above. A post published in the last day or two is still accumulating views, which pulls the 30-day average down slightly. That is a property of the standard definition rather than a fault in it, so we keep the definition rather than “correcting” the number into something nobody can reproduce.

Precision. Telegram publishes view counts on its public widget in short form — 8.12K, 3.7M — so any reading at or above 1,000 reaches us rounded to three significant figures, and only counts below 1,000 are exact. Averages and rates derived from them are shown to the same precision rather than to the unit: a figure like 3,701,250 would assert digits nobody measured.

Reaction counts are published per emoji and rounded the same way, so a total below 1,000 is exact and a larger one is a sum that may carry a rounded component from each emoji above 1,000. Because it is a sum, it does not look rounded — read a large reaction total as three significant figures per contributing emoji rather than as the figure it prints.

Reaction mix

115 reactions across 18 posts, in 3 distinct kinds. The most used accounts for 57.4% of them.

Every reaction kind recorded on the sample, most used first
ReactionCountShareShare, drawn
🔥6657.4%
3127.0%
👍1815.7%

No sentiment is inferred, and none should be read in. This table is ordered by count and by nothing else. Emoji do not carry stable meaning across languages or communities — 🙏 is thanks in one channel and mourning in another — so we publish which ones were pressed and how often, and pass no judgement on what an audience meant by them.

Precision. Telegram publishes reaction counts per emoji and short-forms each one — 4.34K, 1.2M — so any single kind at or above 1,000 reaches us at three significant figures, and only counts below 1,000 are exact. The shares above are ratios of those figures and carry the same error. This is also why the total here can differ slightly from a reaction total printed elsewhere on the page: both are sums of the same rounded parts, taken over samples with different edges.

Coverage. Reactions were read on 18 of the 18 sampled posts in this sample. Summed by Telegram’s own count on each post — not by adding up the per-emoji breakdown above — those same posts carry 115 reactions in total: the kind of figure the paragraph above means by “a reaction total printed elsewhere on the page”.

Measured over the 18 most recent posts we hold, published 7 July 2026 to 7 August 2026, using the newest reading held for each. Telegram Stars are excluded: they are a payment, not a reaction, and they have their own section.

Recent posts

7 Aug 2026, 08:14 UTC97 views1 reactionsread 8 August 2026

За последнюю неделю встретил несколько запросов от топ менеджеров разных компаний на менторинг по Claude Code. Не курсы и лекции, а именно человек который сядет рядом и поможет вместе разобраться на конкретном проекте. Кто-нибудь занимался таким менторингом или есть желание попробовать?

🔥1

5 Aug 2026, 19:49 UTC136 views3 reactionsread 8 August 2026
Photo

В сети уже шутят над слайдами из дека с достижениями команды «А им нужен питч дек» «Мы те, кто создали интернет» «Будут ли VC спрашивать почему эта команда лучшая для работы над стартапом» Но список достижений действительно внушительный - посмотрите сами

1👍1🔥1

5 Aug 2026, 19:37 UTC132 views6 reactionsread 8 August 2026
Photo

Jeff Dean спустя 27 лет работы покидает Google. Наверное один из самых влиятельный и плодотворных инженеров в мире. За его плечам такие технологии как MapReduce, Bigtable, Protobuf. Далее со основал Google Brain и был главным идейным вдохновителем и архитектором Tensorflow. В последнее время руководил исследованиями ИИ и разработкой Gemini. Jeff ушел делать свой стартап - Discovery Loop. Ушел не один, прихватил

🔥6

5 Aug 2026, 05:21 UTC171 views6 reactionsread 8 August 2026
Photo

MoonshotAI запустили AI native credit card. Тратишь деньги - получаешь токены кэшбеком. В добавок получаешь расширенные лимиты на Kimi Code. Пока доступна только в Китае. А вы до сих пор получаете ягодки и мили? ;)

🔥6

3 Aug 2026, 18:24 UTC156 views5 reactionsread 8 August 2026
Photo

Две крутые находки от команды Interconnect, которые помогут мониторить open source модели. The Artifacts Hub — навигатор по Hugging Face. Обзор 792 ключевые модели, выпущенные за последние 2 года. AI index, популярность инференса на openrouter, сравнение моделей друг с другом. Adoption Dashboard — Интерактивная карта скачиваний по странам и организациям. Наглядный способ увидеть разрыв между США и Китаем. Кто из но

5

3 Aug 2026, 05:27 UTC161 views5 reactionsread 8 August 2026
Photo

В 2020 году Giambattista Parascandolo, который сейчас руководит разработкой reasoning моделей в OpenAI, приехал в MIT на собеседование на должность профессора. Вся его презентация была посвящена reasoning моделям. Но большинство профессоров из комиссии лишь посмеялись над этой идеей и сочли ее полной ерундой… По крайней мере, так он написал в биографии на своем сайте ;)

🔥5

1 Aug 2026, 10:14 UTC194 views3 reactionsread 8 August 2026
Photo

В соцсетях заметил мем про меритократию (когда человека оценивают по знаниям и способностям, а не по статусу и богатству) в Сан-Франциско. Мол в СФ оценивают всех как равно только если ты ex. Harvard, ex. MIT, ex. Anthropic, ex. OpenAI. В саммари обсуждения от Grok увидел идеому - Potemkin Village. Пошел искать значение. Потёмкинские деревни Potemkin village) — идиома, означающая показной фасад, построенный, чтобы

🔥3

29 Jul 2026, 18:21 UTC236 views10 reactionsread 8 August 2026

Во времена универа, помню, готовишься к экзамену, а потом незаметно читаешь биографию Геделя или NP полноту. А экзамен по химии… Наблюдаю похожую ситуацию с LLM. Начинаешь делать задачу и из-за простого доступа к любой информации скатываешься в ненужные детали, оптимизации, перфекционизм. У вас наверняка такое бывает тоже. В такие моменты задаю себе два вопроса: Первый: я сейчас реально решают какую-то проблему ил

🔥10

28 Jul 2026, 11:07 UTC247 views8 reactionsread 8 August 2026

Как я вывел для себя новый принцип - Знать нужно ровно столько, чтобы тебя нельзя было обмануть Три вечера делал домашку из CS336. В задании уже есть описанный AGENTS.md который запрещает агенту писать код и сразу выдавать решение. Только подсказки и наставление. Спустя три вечера доделал задание и решил сделать финальное ревью на соответствие требованиям из задания. После нескольких минут раздумий Claude Code выд

🔥71

28 Jul 2026, 05:06 UTC231 views4 reactionsread 8 August 2026

Еще MoonshotAIвыложили свою библиотеку MoonEP - для Expert Parallelism. Вдохновлялись DeepEP от DeepSeek, AcclEP от Alibaba, UltraEP (не понял от какой компании, но авторы из китайских университетов, техрепорт еще). Почему обратил внимание? Сейчас дорешиваю вторую домашку из CS336 - пишу распределенное обучение на многих GPU. Понял, что теперь могу спокойно читать такие статьи и даже понимать. Раньше казалась темны

2🔥2

27 Jul 2026, 20:06 UTC246 views5 reactionsread 8 August 2026

Сегодня выложили веса и технический репорт модели Kimi K3 - самой большой открытой модели. Kimi K3 в 22580 (!) раз больше, чем GPT-2, которая вышла в 2019 году. В твиттере парень решил проследить эволюцию от GPT-2 до Kimi K3. Получилась очень классная статья с примерами кода. Linear Attention Delta Net Gated Delta Net Kimi Linear AttnRes Эти и другие приемы поясняются в статье. По словам автора, на все ушло 48 ч

🔥32

26 Jul 2026, 04:53 UTC248 views10 reactionsread 8 August 2026

Случайно нашел Leetcode для GPU Около 60 задач: начиная от умножения двух векторов до реализации Attention и целых слоев. Почти все задачи являются компонентами нейросетей. Некоторые только недавно встречал в статьях. Например реализация Top K Selection попадалась в Minimax Sparse Attention. Тут у нее стоит уровень hard. Доступны разные языка: PyTorch, Triton, Jax или чистая CUDA. Недавно доделывал домашку по CS3

7🔥3

Showing the 12 most recent of 18 posts we hold for @legchikov_ai. View and reaction counts are the latest single reading for each post, not a live figure, and a recent post is still accumulating both. A view count marked was rounded by Telegram before we ever saw it — t.me prints views in full below 1,000 and to three significant figures above, so ≈1,200,000 means somewhere between 1,150,000 and 1,249,999. Unmarked counts are exact. Text is reproduced from the public post preview and truncated for length.

Citation-graph rank

Citation-graph rank — 1,387,172 of 1,629,535 entries in the measured graph. A weighted position computed from the forward and mention edges below — republished posts weigh more than named mentions — and recomputed periodically, over the whole graph. Published only as this ordinal position, never as a score: a position is a fact, and a score printed beside one channel’s name would read as a verdict this register does not make. The two counts beneath stay separate for the same reason mentions are never summed with forwards anywhere else on this page — a named-by count costs nothing to manufacture. The top 100 by this measure, or how it is computed.

Mentions

Named by 1 registered channel — every channel on the register whose own posts have named this one, by its current username or any other username it currently holds, merged from two separately captured readings of the same fact so a namer caught by only one of them is not missed and a namer both caught is not counted twice. A username this channel has since dropped is not matched — that handle may belong to someone else now, and crediting today’s namer to yesterday’s owner would misattribute it.

Named by

Channels on the register whose posts name this channel's handle.

A mention is a weaker signal than a forward and is counted separately for that reason — naming a channel is not republishing it, and a handle in a post body is easy to place deliberately. The post counts beside each row below are distinct posts in which the handle appeared, from posts we have read on both sides — the “Named by N registered channels” figure above is a different count, of distinct NAMING CHANNELS rather than posts, and is not the sum of the rows under it.

Cite this entry

A live page changes as we take new readings, so a citation should name the measurement it is based on, not just the URL. The line below cites the subscriber count as measured 29 August 2026 — this entry's latest reading, not the date you are reading this.

“• Dmitry Legchikov” (@legchikov_ai), 820 subscribers as measured 29 August 2026. Telegram Register, tgregister.com/channel/legchikov_ai.

Full measurement history, CC BY 4.0. Every reading this register holds for this entry, not just the latest one, as a dated, downloadable record: CSV · JSON. Free to use with attribution to tgregister.com. Each file carries its own generation timestamp, which is the figure to cite for exactly when the data was retrieved.