Channel
@light_from_black_box
On this record: Growth · Engagement · Reactions · Posts · Citations · Cite this entry
1,178subscribers
+6 since we began measuring on 6 August 2026
Risers and fallers across the register · movement among entries of 1,000–3,162.
| Telegram ID | -1002608998064 |
|---|---|
| Type | Channel |
| Username | @light_from_black_box |
| Created | Between 1 March 2025 and 31 July 2025— estimated from Telegram’s id allocation, not measured. How this range is calculated. |
| First recorded | 6 August 2026 |
| Last confirmed live | 13 August 2026 |
| Measurements held | 4 |
| Confirmed unchanged | 1 time, most recently 13 August 2026 |
| On Telegram | t.me/light_from_black_box |
| Measured (UTC) | Subscribers | Change |
|---|---|---|
| 13 Aug 2026, 06:24 | 1,178 | +13 |
| 10 Aug 2026, 09:58 | 1,165 | +2 |
| 7 Aug 2026, 11:59 | 1,163 | -9 |
| 6 Aug 2026, 18:17 | 1,172 | first reading |
18 posts held, back to 12 September 2025 — the reader has not yet reached the start of this channel’s public history, so older posts may sit further back, unread. Read across 1 pageof Telegram’s post history, 20 posts per page.
Nothing published in the last 30 days. ERR and ER are rolling 30-day measures, so there is nothing to compute — we hold 18 posts for this entry, the most recent from 17 March 2026. An engagement rate over an empty window would be a number about nothing.
554 reactions across 18 posts, in 8 distinct kinds. The most used accounts for 66.1% of them.
| Reaction | Count | Share | Share, drawn |
|---|---|---|---|
| 🔥 | 366 | 66.1% | |
| ❤ | 81 | 14.6% | |
| 👍 | 79 | 14.3% | |
| 👀 | 12 | 2.17% | |
| 👏 | 7 | 1.26% | |
| 🙏 | 4 | 0.722% | |
| 🥴 | 3 | 0.542% | |
| 💩 | 2 | 0.361% |
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 554reactions 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 12 September 2025 to 17 March 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.
Читала сегодня в ИТМО лекцию по Semantic IDs и Generative Retrieval в рамках курса по продвинутым рекомендательным системам. Первый опыт, когда материал для лекции я собирала с нуля - на основе большого количества статей по теме. А сами статьи перед этим много и подробно разбирали внутри команды. Было здорово разложить наконец всё по полочкам) Материалом делюсь
🔥17👍12❤4
Свет из чёрного ящика Когда я запускал этот канал, то писал, что буду рассказывать об успехах и неудачах нашей команды. Однако до сих пор речь в основном шла о том, что происходит в индустрии. Настало время поделиться апдейтами 🙂 За последние полгода наша команда полностью перестроилась на R&D. Направлений исследования стало больше, они расширяются. Чем мы занимаемся последнее время: * Адаптировали Argus и Deep R…
🔥31❤7🥴3💩2👍1
Пока у нас новый год и праздники, компании из Китая продолжают выкладывать очень интересные работы, так что выходные можно провести с пользой. * OxygenREC: An Instruction-Following Generative Framework for E-commerce Recommendation JD.com опубликовала работу про внедрение end2end генеративной рекомендательной системы в разные поврехности их маркетплейса. Результаты АБ-тестов впечатляют. Куча технических подробносте…
🔥13
Подборка статей 2025 (часть 3) Attention-based ranking - LONGER: Scaling Up Long Sequence Modeling in Industrial Recommenders - TransAct V2: Lifelong User Action Sequence Modeling on Pinterest Recommendation - Twin-Flow Generative Ranking Network for Recommendation - InterFormer: Effective Heterogeneous Interaction Learning for Click-Through Rate Prediction - MARM: Unlocking the Recommendation Cache Scaling-Law thro…
🔥14👍4👀2
Подборка статей 2025 (часть 2) Фундаментальные модели - PinFM: Foundation Model for User Activity Sequences at a Billion-scale Visual Discovery Platform - Large Foundation Model for Ads Recommendation - Foundation Model for Personalized Recommendation - Meta’s Generative Ads Model (GEM): The Central Brain Accelerating Ads Recommendation AI Innovation - Realizing Scaling Laws in Recommender Systems: A Foundation–Expe…
👍18👀6🔥4❤2
Подборка статей 2025 (часть 1) Как и обещал, выкладываю подборку статей по темам, которые освещал на выступлении. End2End - OneRec: Unifying Retrieve and Rank with Generative Recommender and Iterative Preference Alignment - OneRec Technical Report - OneRec-V2 Technical Report - OneLoc: Geo-Aware Generative Recommender Systems for Local Life Service - OneSug: The Unified End-to-End Generative Framework for E-commerc…
🔥32👍15❤9👀4
File, posted without a caption
🔥32❤2
ML Global Recap'25 Готовлю выступление на наш митап ML Global Recap. Расскажу там по-подробнее, как я вообще вижу текущие тренды в рекомендациях, куда идут разные компании и что рассказали на конференциях в этом году. По-подробнее расскажу, что было на RecSys (я на него не ездил, но от нашей команды была большая делегация) и на CIKM, добавлю каких-нибудь инсайдов. Кроме меня выступит Лёша Гусаков, наш CTO, с расска…
🔥21
CIKM'25 recap Для меня это первый опыт поездки на CIKM и надо сказать, что конференция показалась мне в разы более полезной, чем RecSys. У обеих нет статуса A*, но CIKM отличается более лайтовым ревью работ. Например там нет ограничений на публикацию статьи до самой конфы. Казалось бы, такой подход должен привести к негативному отбору, но по факту на конференцию привозят большое количество отличных индустриальных и …
👍16🔥8🙏4❤3
CIKM'25 В этом году я решил не ехать на RecSys, зато лечу на CIKM в Сеул. Не все слышали про эту конференцию, хотя на ней публикуются очень многие хорошие работы. В программе этого года самые интересные (на мой взгляд) работы от Pinterest (TransAct V2 и PinRec), несколько работ от команды OneRec из Kuaishou (например QARM про способ токенизации и Pantheon, который они используют в качестве reward модели), а также e…
🔥29❤17
Yandex Cup Пока я немного погряз в работе и запусках (надеюсь очень скоро будет, чем поделиться в канал), напишу здесь про Yandex Cup. В этом году пересобрали ML трек, задачи должны стать интереснее, а финал пройдет в Стамбуле. Если интересно попробовать свои силы, еще остается несколько дней на регистрацию: https://yandex.ru/cup
❤12🔥6👍3
Презентация с сегодняшнего выступления Всем спасибо, кто дошел, было приятно видеть полный зал!
🔥34❤9👏7
Showing the 12 most recent of 18 posts we hold for @light_from_black_box. 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 — 7,008 of 1,350,102entries 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.
Republished by
Channels on the register that have forwarded this channel's posts into their own feed.
Republishes
Channels on the register whose posts this channel has forwarded.
Built only from forwarded posts we have actually read, on both sides. Coverage is early and deliberately incomplete: a missing link means we have not read the post that would prove it, never that the relationship does not exist. Counts are distinct forwarded posts observed, so they only ever go up as we read more.
Named by 6 registered channels — 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.
Names
Channels on the register whose handles appear in this channel's posts.
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.
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 13 August 2026 — this entry's latest reading, not the date you are reading this.
“Свет из чёрного ящика” (@light_from_black_box), 1,178 subscribers as measured 13 August 2026. Telegram Register, tgregister.com/channel/light_from_black_box.
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.