task Boris-1.7 BM-2-Medium ฮ HellaSwag 31.22 32.43 โ1.21 ARC-Easy 45.88 43.81 +2.07 ARC-Challenge 26.02 25.34 +0.68 PIQA 62.57 61.86 +0.71 4-task average 41.42 40.86 +0.56
Arithmark 3? and bananamind bench?
๐ Vega 1.6 128M elo going up
CNWPlayer
CNWPlayer
AI & ML interests
SLMs, LLMs, really anything that ends in LM.
Recent Activity
repliedto KlondikeDev's post about 14 hours ago
A Preview of Boris-2!
Hello! Tomorrow, OpenCerebral will be releasing Boris-1.7-D60M-n30M โ an experimental architecture. It will be testing a new data mixture, a new tokenizer, and testing Qwen4-like n-gram embeddings.
Following this will be Boris-1.8-D60M-n30M, which will test both the n-gram embeddings AND a new architecture.
Then, Boris-2 will begin training!
Edit: the model is OUT NOW! https://huggingface.co/opencerebral/Boris-1.7-D60M-n30M repliedto KlondikeDev's post 1 day ago
A Preview of Boris-2!
Hello! Tomorrow, OpenCerebral will be releasing Boris-1.7-D60M-n30M โ an experimental architecture. It will be testing a new data mixture, a new tokenizer, and testing Qwen4-like n-gram embeddings.
Following this will be Boris-1.8-D60M-n30M, which will test both the n-gram embeddings AND a new architecture.
Then, Boris-2 will begin training!
Edit: the model is OUT NOW! https://huggingface.co/opencerebral/Boris-1.7-D60M-n30M reacted to KlondikeDev's post with ๐ฅ 1 day ago
A Preview of Boris-2!
Hello! Tomorrow, OpenCerebral will be releasing Boris-1.7-D60M-n30M โ an experimental architecture. It will be testing a new data mixture, a new tokenizer, and testing Qwen4-like n-gram embeddings.
Following this will be Boris-1.8-D60M-n30M, which will test both the n-gram embeddings AND a new architecture.
Then, Boris-2 will begin training!
Edit: the model is OUT NOW! https://huggingface.co/opencerebral/Boris-1.7-D60M-n30MOrganizations
None yet
replied to KlondikeDev's post about 14 hours ago
replied to KlondikeDev's post 1 day ago
will be interesting to see the effects of this
reacted to KlondikeDev's post with ๐ฅ 1 day ago
Post
2673
A Preview of Boris-2!
Hello! Tomorrow, OpenCerebral will be releasing Boris-1.7-D60M-n30M โ an experimental architecture. It will be testing a new data mixture, a new tokenizer, and testing Qwen4-like n-gram embeddings.
Following this will be Boris-1.8-D60M-n30M, which will test both the n-gram embeddings AND a new architecture.
Then, Boris-2 will begin training!
Edit: the model is OUT NOW! opencerebral/Boris-1.7-D60M-n30M
Hello! Tomorrow, OpenCerebral will be releasing Boris-1.7-D60M-n30M โ an experimental architecture. It will be testing a new data mixture, a new tokenizer, and testing Qwen4-like n-gram embeddings.
Following this will be Boris-1.8-D60M-n30M, which will test both the n-gram embeddings AND a new architecture.
Then, Boris-2 will begin training!
Edit: the model is OUT NOW! opencerebral/Boris-1.7-D60M-n30M
- 53 replies
liked a
Space 1 day ago
reacted to Banaxi-Tech's post with ๐ 1 day ago
Post
3463
We're delaying BananaMind 2.1!
When BananaMind 2.1 Lite was almost done, we benchmarked it and the results we're worse than BananaMind 2 Mini.
We're going to spend alot more time in research on tiny models and then scaling up our techniques to the actual BananaMind 2.1 models!
We're also announcing these new models:
BananaMind 2.1 Coder: A 149M instruction tuned coder model trained on 75B tokens + 10B tokens of stack-v3-train.
BananaMind 2.1 Pico: A 1M parameter model trained on 22B tokens of data.
We also may release BananaMind 2.1 Large with around 100M parameters depending on how much compute we have.
Please give us a follow!
BananaMind
@Banaxi-Tech
---
@vovaRL
@DedeProGames
When BananaMind 2.1 Lite was almost done, we benchmarked it and the results we're worse than BananaMind 2 Mini.
We're going to spend alot more time in research on tiny models and then scaling up our techniques to the actual BananaMind 2.1 models!
We're also announcing these new models:
BananaMind 2.1 Coder: A 149M instruction tuned coder model trained on 75B tokens + 10B tokens of stack-v3-train.
BananaMind 2.1 Pico: A 1M parameter model trained on 22B tokens of data.
We also may release BananaMind 2.1 Large with around 100M parameters depending on how much compute we have.
Please give us a follow!
@Banaxi-Tech
---
@vovaRL
@DedeProGames
- 7 replies
Update numbers
11
#29 opened 2 days ago
by CNWPlayer
Would be nice to have raw Elos available with some sort of switch
1
#28 opened 3 days ago
by CNWPlayer
updated 2
models 2 days ago
reacted to Bc-AI's post with ๐๐ฅ 2 days ago
Post
2008
Hello everyone! Happy to say that MiniCoder-1 is now in the instruction tuning phase. It is a 216M~ parameter model trained on 16B tokens. It ran on an RTX 6000 Pro Blackwell gpu for around 24~ hours. Now we will do SFT and launch as beta while we work on the final important DPO and RLHF phases. Our goal is a small extreemly fast on device coding assistant with CoT reasoning* baked in! - Bc-AI on behalf of the Smilyai-Labs team
*it is a small model so the reasoning quality wont be as good obviously!
*it is a small model so the reasoning quality wont be as good obviously!
- 2 replies
published a
model 2 days ago
updated a
model 3 days ago
Add VegaLM1-42M-Base and VegaLM1-42M-CodeCompletion
1
#16 opened 6 days ago
by CNWPlayer
Add VegaLM1-42M-Base
1
#108 opened 6 days ago
by CNWPlayer
๐ฉ Report: Spam
๐โ 2
1
#2 opened 7 days ago
by CNWPlayer
reacted to Banaxi-Tech's post with ๐๐ฅ 8 days ago
Post
2698
We're releasing Overfitter 1.0.
Its a completly useless overfitted model trained on 202 epochs of SWE Bench Verified, SWE Bench Pro, Terminal Bench 2.1, DeepSWE.
It gets 100% on SWE Bench Verified, 98.6% on SWE bench Pro, 100% on Terminal Bench 2.1 and 100% on DeepSWE!
BananaMind/Overfitter-1.0
Its a completly useless overfitted model trained on 202 epochs of SWE Bench Verified, SWE Bench Pro, Terminal Bench 2.1, DeepSWE.
It gets 100% on SWE Bench Verified, 98.6% on SWE bench Pro, 100% on Terminal Bench 2.1 and 100% on DeepSWE!
BananaMind/Overfitter-1.0
- 11 replies
published a
model 9 days ago
reacted to dejanseo's post with ๐ฅ 9 days ago
Post
2991
Ox Alpha is GLM
https://dejan.ai/blog/ox-alpha/
A parameter-free k-nearest-neighbour classifier over Normalized Compression Distance (Lee et al., MobiSys โ24, Eq. 1, built on Jiang et al.'s gzip-based text classifier). NCD compares two texts by how well they compress together. C(s) is the gzip-compressed length of s. Text sharing an author's patterns compresses better together than text from a different author, so the method needs no model weights and no embeddings.
The reference corpus covers 60 prompts (essays, code, emails, dialogue, poetry) answered by five known models: GPT-5.5, Claude Opus 5, Gemini 3.7 Flash, Gemini 3.1 Pro Preview, and GLM-5.3, for 293 reference texts. ox-alpha answered the first 13 of those prompts, plus one additional novel prompt never given to the reference models beforehand, for 14 queries in total. Each query was classified against the reference corpus independently, with a k-nearest-neighbour vote (k=5):
Model ox-alpha samples matched
GLM-5.3 7 / 14
Claude Opus 5 3 / 14
Gemini 3.7 Flash 2 / 14
GPT-5.5 1 / 14
Gemini 3.1 Pro Preview 1 / 14
GLM-5.3 wins at every k tested: 7/14 at k=3, 7/14 at k=5, 6/14 at k=7, 7/14 at k=9. Claude Opus 5 is the consistent second place.
https://dejan.ai/blog/ox-alpha/
A parameter-free k-nearest-neighbour classifier over Normalized Compression Distance (Lee et al., MobiSys โ24, Eq. 1, built on Jiang et al.'s gzip-based text classifier). NCD compares two texts by how well they compress together. C(s) is the gzip-compressed length of s. Text sharing an author's patterns compresses better together than text from a different author, so the method needs no model weights and no embeddings.
The reference corpus covers 60 prompts (essays, code, emails, dialogue, poetry) answered by five known models: GPT-5.5, Claude Opus 5, Gemini 3.7 Flash, Gemini 3.1 Pro Preview, and GLM-5.3, for 293 reference texts. ox-alpha answered the first 13 of those prompts, plus one additional novel prompt never given to the reference models beforehand, for 14 queries in total. Each query was classified against the reference corpus independently, with a k-nearest-neighbour vote (k=5):
Model ox-alpha samples matched
GLM-5.3 7 / 14
Claude Opus 5 3 / 14
Gemini 3.7 Flash 2 / 14
GPT-5.5 1 / 14
Gemini 3.1 Pro Preview 1 / 14
GLM-5.3 wins at every k tested: 7/14 at k=3, 7/14 at k=5, 6/14 at k=7, 7/14 at k=9. Claude Opus 5 is the consistent second place.
- 6 replies