crownelius/Opus-4.6-Reasoning-3300x
Viewer
β’ Updated
β’ 2.16k β’ 1.39k β’ 129
Mex Ivanov
MexIvanov
AI & ML interests
NLP, Coding, Quantum Computing and more.
Recent Activity
liked
a dataset 5 days ago
crownelius/Opus-4.6-Reasoning-3300x reacted
to
marksverdhei's
post with π 28 days ago
Poll: Will 2026 be the year of subquadratic attention?
The transformer architecture is cursed by its computational complexity.
It is why you run out of tokens and have to compact. But some would argue that this is a feature not a bug and that this is also why these models are so good. We've been doing a lot of research on trying to make equally good models that are computationally cheaper, But so far, none of the approaches have stood the test of time. Or so it seems.
Please vote, don't be shy. Remember that the Dunning-Kruger effect is very real, so the person who knows less about transformers than you is going to vote. We want everyone's opinion, no matter confidence.
π if you think at least one frontier model* will have no O(n^2) attention by the end of 2026
π₯ If you disagree
* Frontier models - models that match / outperform the flagship claude, gemini or chatgpt at the time on multiple popular benchmarks liked
a model about 2 months ago
google/translategemma-27b-it Organizations
None yet
liked a dataset 5 days ago
reacted to marksverdhei's post with π 28 days ago
Post
4574
Poll: Will 2026 be the year of subquadratic attention?
The transformer architecture is cursed by its computational complexity.
It is why you run out of tokens and have to compact. But some would argue that this is a feature not a bug and that this is also why these models are so good. We've been doing a lot of research on trying to make equally good models that are computationally cheaper, But so far, none of the approaches have stood the test of time. Or so it seems.
Please vote, don't be shy. Remember that the Dunning-Kruger effect is very real, so the person who knows less about transformers than you is going to vote. We want everyone's opinion, no matter confidence.
π if you think at least one frontier model* will have no O(n^2) attention by the end of 2026
π₯ If you disagree
* Frontier models - models that match / outperform the flagship claude, gemini or chatgpt at the time on multiple popular benchmarks
The transformer architecture is cursed by its computational complexity.
It is why you run out of tokens and have to compact. But some would argue that this is a feature not a bug and that this is also why these models are so good. We've been doing a lot of research on trying to make equally good models that are computationally cheaper, But so far, none of the approaches have stood the test of time. Or so it seems.
Please vote, don't be shy. Remember that the Dunning-Kruger effect is very real, so the person who knows less about transformers than you is going to vote. We want everyone's opinion, no matter confidence.
π if you think at least one frontier model* will have no O(n^2) attention by the end of 2026
π₯ If you disagree
* Frontier models - models that match / outperform the flagship claude, gemini or chatgpt at the time on multiple popular benchmarks
- 4 replies
liked 3 models about 2 months ago
reacted to RakshitAralimatti's post with π₯ 3 months ago
Post
2456
I built something crazy you never saw before.
Please check - https://huggingface.co/blog/RakshitAralimatti/streaming-data-rag
A real-time Streaming Data to RAG system that listens to live radio, transcribes it on-the-fly, and lets you query across TIME.
Not just "what was discussed" β but "what happened in the last 10 minutes on channel 0?" or "at 9 AM, what was the breaking news?" This is RAG that understands temporal context.
Please check - https://huggingface.co/blog/RakshitAralimatti/streaming-data-rag
A real-time Streaming Data to RAG system that listens to live radio, transcribes it on-the-fly, and lets you query across TIME.
Not just "what was discussed" β but "what happened in the last 10 minutes on channel 0?" or "at 9 AM, what was the breaking news?" This is RAG that understands temporal context.
- 1 reply
reacted to samerzaher80's post with π 3 months ago
Post
1629
AetherMind_SRL: How I beat 7B models on MMLU with 184M params and a $300 GPU
Iβm Sameer, a solo researcher from Iraq working on a single RTX 3050 8GB laptop.Today Iβm releasing AetherMind_SRL β a 184M-parameter NLI model that was trained only on tasks (SNLI, MNLI, ANLI, and a small clinical Alzheimerβs dataset).
It was never fine-tuned or even shown a single MMLU question during training.Yet here are the zero-shot MMLU (57 subjects) results:Model
MMLU Zero-Shot
Training Data
AetherMind_SRL (me)
184M
36.05 %
Only NLI (SNLI/MNLI/ANLI + ADNI)
DeBERTa-v3-base
278M
~30.8 %
General pre-training
BERT-large
340M
27β30 %
General pre-training
LLaMA-1 7B
7B
34β35 %
Massive text corpus
LLaMA-2 7B
7B
~45 %
Bigger + better data
Yes β my 184M model beats every classic 300β400M model and the original 7-billion-parameter LLaMA-1, all while running at 300+ samples/sec on a $300 laptop GPU.How did this happen?I built a standardized self-improvement loop called AetherMind Self-Reflective Learning (SRL) v1.0:Train normally on NLI
Let the model predict on hard adversarial data (ANLI)
Log every mistake + low-confidence case
Build a balanced βSMARTβ buffer (60% errors + 40% correct anchors)
Fine-tune with tiny LR and error-weighted loss
Repeat until stable
Thatβs it. No external knowledge, no MMLU data, no cluster.
Just pure reasoning transfer from entailment/contradiction patterns β real-world knowledge.Try it yourself python
from transformers import pipeline
import torch
nli_pipeline = pipeline(
"text-classification",
model="samerzaher80/AetherMind_SRL",
device=0 if torch.cuda.is_available() else -1
)
# DEFINE YOUR TEST HERE
premise = "Patient shows progressive memory decline."
hypothesis = "Patient shows progressive memory decline."
input_text = f"{premise} [SEP] {hypothesis}"
result = nli_pipeline(input_text)[0]
print(f"Prediction: {result['label']}")
print(f"Confidence: {result['score']:
Model: samerzaher80/AetherMind_SRL
Iβm Sameer, a solo researcher from Iraq working on a single RTX 3050 8GB laptop.Today Iβm releasing AetherMind_SRL β a 184M-parameter NLI model that was trained only on tasks (SNLI, MNLI, ANLI, and a small clinical Alzheimerβs dataset).
It was never fine-tuned or even shown a single MMLU question during training.Yet here are the zero-shot MMLU (57 subjects) results:Model
MMLU Zero-Shot
Training Data
AetherMind_SRL (me)
184M
36.05 %
Only NLI (SNLI/MNLI/ANLI + ADNI)
DeBERTa-v3-base
278M
~30.8 %
General pre-training
BERT-large
340M
27β30 %
General pre-training
LLaMA-1 7B
7B
34β35 %
Massive text corpus
LLaMA-2 7B
7B
~45 %
Bigger + better data
Yes β my 184M model beats every classic 300β400M model and the original 7-billion-parameter LLaMA-1, all while running at 300+ samples/sec on a $300 laptop GPU.How did this happen?I built a standardized self-improvement loop called AetherMind Self-Reflective Learning (SRL) v1.0:Train normally on NLI
Let the model predict on hard adversarial data (ANLI)
Log every mistake + low-confidence case
Build a balanced βSMARTβ buffer (60% errors + 40% correct anchors)
Fine-tune with tiny LR and error-weighted loss
Repeat until stable
Thatβs it. No external knowledge, no MMLU data, no cluster.
Just pure reasoning transfer from entailment/contradiction patterns β real-world knowledge.Try it yourself python
from transformers import pipeline
import torch
nli_pipeline = pipeline(
"text-classification",
model="samerzaher80/AetherMind_SRL",
device=0 if torch.cuda.is_available() else -1
)
# DEFINE YOUR TEST HERE
premise = "Patient shows progressive memory decline."
hypothesis = "Patient shows progressive memory decline."
input_text = f"{premise} [SEP] {hypothesis}"
result = nli_pipeline(input_text)[0]
print(f"Prediction: {result['label']}")
print(f"Confidence: {result['score']:
Model: samerzaher80/AetherMind_SRL
liked 4 models 4 months ago
reacted to tomaarsen's post with π€π 5 months ago
Post
4514
π€ Sentence Transformers is joining Hugging Face! π€ This formalizes the existing maintenance structure, as I've personally led the project for the past two years on behalf of Hugging Face! Details:
Today, the Ubiquitous Knowledge Processing (UKP) Lab is transferring the project to Hugging Face. Sentence Transformers will remain a community-driven, open-source project, with the same open-source license (Apache 2.0) as before. Contributions from researchers, developers, and enthusiasts are welcome and encouraged. The project will continue to prioritize transparency, collaboration, and broad accessibility.
Read our full announcement for more details and quotes from UKP and Hugging Face leadership: https://huggingface.co/blog/sentence-transformers-joins-hf
We see an increasing wish from companies to move from large LLM APIs to local models for better control and privacy, reflected in the library's growth: in just the last 30 days, Sentence Transformer models have been downloaded >270 million times, second only to transformers.
I would like to thank the UKP Lab, and especially Nils Reimers and Iryna Gurevych, both for their dedication to the project and for their trust in myself, both now and two years ago. Back then, neither of you knew me well, yet you trusted me to take the project to new heights. That choice ended up being very valuable for the embedding & Information Retrieval community, and I think this choice of granting Hugging Face stewardship will be similarly successful.
I'm very excited about the future of the project, and for the world of embeddings and retrieval at large!
Today, the Ubiquitous Knowledge Processing (UKP) Lab is transferring the project to Hugging Face. Sentence Transformers will remain a community-driven, open-source project, with the same open-source license (Apache 2.0) as before. Contributions from researchers, developers, and enthusiasts are welcome and encouraged. The project will continue to prioritize transparency, collaboration, and broad accessibility.
Read our full announcement for more details and quotes from UKP and Hugging Face leadership: https://huggingface.co/blog/sentence-transformers-joins-hf
We see an increasing wish from companies to move from large LLM APIs to local models for better control and privacy, reflected in the library's growth: in just the last 30 days, Sentence Transformer models have been downloaded >270 million times, second only to transformers.
I would like to thank the UKP Lab, and especially Nils Reimers and Iryna Gurevych, both for their dedication to the project and for their trust in myself, both now and two years ago. Back then, neither of you knew me well, yet you trusted me to take the project to new heights. That choice ended up being very valuable for the embedding & Information Retrieval community, and I think this choice of granting Hugging Face stewardship will be similarly successful.
I'm very excited about the future of the project, and for the world of embeddings and retrieval at large!
- 1 reply
liked a model 5 months ago
liked a Space 5 months ago
liked a model 6 months ago
liked a dataset 10 months ago
liked a model 11 months ago
reacted to takarajordan's post with π 12 months ago
Post
1881
Takara takes 3rd place in the {tech:munich} AI hackathon with Fudeno!
A little over 2 weeks ago @aldigobbler and I set out to create the largest MultiModal SVG dataset ever created, we succeeded in this and when I was in Munich, Germany I took it one step further and made an entire app with it!
We fine-tuned Mistral Small, made a Next.JS application and blew some minds, taking 3rd place out of over 100 hackers. So cool!
If you want to see the dataset, please see below.
takara-ai/fudeno-instruct-4M
A little over 2 weeks ago @aldigobbler and I set out to create the largest MultiModal SVG dataset ever created, we succeeded in this and when I was in Munich, Germany I took it one step further and made an entire app with it!
We fine-tuned Mistral Small, made a Next.JS application and blew some minds, taking 3rd place out of over 100 hackers. So cool!
If you want to see the dataset, please see below.
takara-ai/fudeno-instruct-4M
liked a model 12 months ago
