---
title: "If you train models with MTP architecture, stop injecting noisy CE gradients and instead add the… — מייק — DecisionGateway"
description: "commitment by מייק, recorded recently — verbatim quote and source citation."
canonical_url: "https://dgw.tmarket.io/claims/01a1e1ee-4898-7ced-9f43-9f192a44b50c"
meta:
  "og:description": "commitment by מייק, recorded recently — verbatim quote and source citation."
  "og:image": "https://dgw.tmarket.io/og.jpg"
  "og:image:alt": "White network of connected nodes crossed by a timeline on black"
  "og:image:height": "1024"
  "og:image:width": "1024"
  "og:title": "If you train models with MTP architecture, stop injecting noisy CE gradients and instead add the… — מייק — DecisionGateway"
  "twitter:image": "https://dgw.tmarket.io/og.jpg"
---

> Index of every page on this site: [llms.txt](https://dgw.tmarket.io/llms.txt)

commitment

# **If you train models with MTP architecture, stop injecting noisy CE gradients and instead add the Soft-MTP line of code.**

A commitment by [**מייק**](https://dgw.tmarket.io/actors/01a1e1eb-e2e0-7491-988c-4db9047ab762). Statements are as reported by the cited publishers — not verified by DecisionGateway.

> *אם אתם מאמנים מודלים עם ארכיטקטורת MTP, תפסיקו להזריק גראדיאנטים רועשים של CE אלא הוסיפו את שורת הקוד של Soft-MTP*

mathyaiwithmikePublished 2026-11-05 [ Read source](https://t.me/mathyaiwithmike/1303)

[**מייק dossier **](https://dgw.tmarket.io/actors/01a1e1eb-e2e0-7491-988c-4db9047ab762)

## **Discussion**

## **Related claims**

[**Soft-MTP obtains the same gradient for free by a simple subtraction between gradients already computed, instead of computing KL from scratch (which requires a lot of memory).**post author · statement](https://dgw.tmarket.io/claims/01a1e1ee-ca55-7120-bcef-603814df663d) [**When decomposing the standard MTP gradient (trained with standard Cross-Entropy), two opposing forces are hidden inside: a useful self-distillation signal and a residual noise component.**post author · statement](https://dgw.tmarket.io/claims/01a1e1ee-c9b5-7ec8-958e-20cbda9fe423) [**Large AI labs train models with MTP, but 80% of the gradient they backpropagate to the model is static noise that interferes with learning.**post author · allegation](https://dgw.tmarket.io/claims/01a1e1ee-c8cc-76f8-a0cf-fdc0d5c6fade) [**Adding the Soft-MTP line of code isolates the Self-Distillation signal and yields a smarter, faster model in speculative decoding without paying an additional penny in compute or memory.**מייק · statement](https://dgw.tmarket.io/claims/01a1e1ee-48b9-76c9-ba93-6ab9598c88a7) [**MTP not only improves performance in math and code, but also provides built-in draft heads for inference acceleration via Speculative Decoding.**post author · statement](https://dgw.tmarket.io/claims/01a1e1ee-c8fd-7400-9123-603ba0e22ca1) [**In the Effective Rank metric, regular MTP training collapsed the output matrix from 496 to 131, while Soft-MTP kept it high and stable at 437.**post author · statement](https://dgw.tmarket.io/claims/01a1e1ee-cad6-7a16-a096-11c8854a7260)

## **More claims by מייק**

[**In event-driven neuromorphic architectures, parts of the network that do not need to work simply stay at rest.**statement](https://dgw.tmarket.io/claims/01a1e1f5-cf7d-742b-bd28-7eafcdfec5f9) [**In event-driven neuromorphic architectures, memory is adjacent to the processor.**statement](https://dgw.tmarket.io/claims/01a1e1f5-cf6f-7b10-a7d1-2224693c7c3b) [**In event-driven neuromorphic architectures, time is a computational feature.**statement](https://dgw.tmarket.io/claims/01a1e1f5-cf5d-7f20-8538-3cbeded37e52) [**The future of AI lies not only in pushing more billions of parameters into hot GPUs, but in a smart combination of classic transformers and LLMs with event-driven neuromorphic architectures.**prediction](https://dgw.tmarket.io/claims/01a1e1f5-cf34-7457-8dce-82a66f8c276f) [**EBMs promised a revolution 20 years ago, and the problem was never elegance — it was training at scale.**statement](https://dgw.tmarket.io/claims/01a1e1f3-ceae-7241-9d49-f814ac4ed3e0)

[All claims by מייק](https://dgw.tmarket.io/claims?claimant_id=01a1e1eb-e2e0-7491-988c-4db9047ab762)

[All claims](https://dgw.tmarket.io/claims)

```json
{"@context":"https://schema.org","@graph":[{"@id":"https://dgw.tmarket.io/#website","@type":"WebSite","description":"Decisions, events and falsifiable forecasts from 4,500+ Telegram channels — every claim cited. Read free.","name":"DecisionGateway","url":"https://dgw.tmarket.io/","publisher":{"@id":"https://dgw.tmarket.io/#identity"}},{"@id":"https://dgw.tmarket.io/claims/01a1e1ee-4898-7ced-9f43-9f192a44b50c#webpage","@type":"WebPage","description":"commitment by מייק, recorded recently — verbatim quote and source citation.","name":"If you train models with MTP architecture, stop injecting noisy CE gradients and instead add the… — מייק — DecisionGateway","url":"https://dgw.tmarket.io/claims/01a1e1ee-4898-7ced-9f43-9f192a44b50c","breadcrumb":{"@id":"https://dgw.tmarket.io/claims/01a1e1ee-4898-7ced-9f43-9f192a44b50c#breadcrumb"},"isPartOf":{"@id":"https://dgw.tmarket.io/#website"},"potentialAction":[{"@type":"ReadAction","target":["https://dgw.tmarket.io/claims/01a1e1ee-4898-7ced-9f43-9f192a44b50c"]}]},{"@id":"https://dgw.tmarket.io/#identity","@type":"Organization","name":"DecisionGateway","url":"https://dgw.tmarket.io"},{"@id":"https://dgw.tmarket.io/claims/01a1e1ee-4898-7ced-9f43-9f192a44b50c#breadcrumb","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","name":"Home","item":"https://dgw.tmarket.io","position":1},{"@type":"ListItem","name":"Claims","item":"https://dgw.tmarket.io/claims","position":2},{"@type":"ListItem","name":"If you train models with MTP architecture, stop injecting…","position":3}]},{"@id":"https://dgw.tmarket.io/claims/01a1e1ee-4898-7ced-9f43-9f192a44b50c#/schema/quotation/1","@type":"Quotation","dateCreated":"2026-11-05","isBasedOn":"https://t.me/mathyaiwithmike/1303","text":"אם אתם מאמנים מודלים עם ארכיטקטורת MTP, תפסיקו להזריק גראדיאנטים רועשים של CE אלא הוסיפו את שורת הקוד של Soft-MTP","url":"/claims/01a1e1ee-4898-7ced-9f43-9f192a44b50c","spokenByCharacter":{"@type":"Person","name":"מייק","@id":"/actors/01a1e1eb-e2e0-7491-988c-4db9047ab762"}}]}
```

## Sitemap

See the full [sitemap](https://dgw.tmarket.io/sitemap.md) for all pages.
