# Symbiotic-Beta by Convergent Intelligence: Open-Weight Model Source: https://savrn.com/models/symbiotic-beta Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve Symbiotic-Beta (3.6B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them. | Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits | | --- | --- | --- | --- | --- | --- | | 16-bit | 7.1 GB | 8.6 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 8-bit | 3.6 GB | 4.3 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 4-bit | 1.8 GB | 2.1 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the [SAVRN Index](https://savrn.com/ai-index/pricing/gpus), read Oct 25, 2026. [Symbiotic-Beta on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/symbiotic-beta/gpus) ## Model Card By Convergent Intelligence, published under afl-3.0, revision 64a0a5b6320c. ### SymLM SymbioticLM is a hybrid symbolic–neural language model that integrates a frozen transformer backbone (Qwen2ForCausalLM) with a suite of symbolic cognitive modules for adaptive, interpretable reasoning. ### Model Description The architecture fuses neural token-level generation with symbolic introspection and reasoning: - Dynamic Thought Evolution with Helical Encoding and DNA-Inspired Memory (DTE-HDM) Enables structured long-term memory and spiral-context encoding across tokens. - Multi-Agent Symbiotic Response Mechanisms (M.A.S.R.M) Coordinates symbolic-neural agents via gated attention and adaptive response layers. - QwenExoCortex Projects contextual hidden states from the Qwen model into a symbolic fusion space for reasoning and memory replay. - Symbolic processors Includes: - ThoughtDynamicsLNN - Liquid / Crystalline Processors - Graph Reasoning with DNAConv - A rolling ThoughtMemory This enables real-time fusion of symbolic thinking, token generation, and reasoning-aware language modeling. ### Intended Uses & Limitations #### Intended Uses [Read the full model card (794 words)](https://savrn.com/models/symbiotic-beta/card) ## Configuration Architecture Qwen2ForCausalLM Hidden size 2,048 Feed-forward size 8,192 Attention heads 32 Key/value heads 32 Vocabulary size 37,524 Stored precision float16 Model type symbiotic-llm ## Identity and Version Repository reaperdoesntknow/Symbiotic-Beta Publisher Convergent Intelligence Task Text generation Modality Text Library transformers Parameters 3.6B parameters Languages en Revision 64a0a5b6320ca465b8378ecec07903764f4744d6 First published 2025-04-26 Last updated 2026-09-18 ## Files and Weights 11 files, 35.9 GB in total. The weights are 4 files totalling 35.9 GB in bin, pt, safetensors. Weights4 files · 35.9 GB Configuration3 files · 2.2 KB Tokenizer2 files · 4.5 MB Documentation1 file · 8.7 KB Repository1 file · 1.5 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | model.safetensors | Weights | 14.3 GB | f7bb78bf0f1b | | pytorch_model.bin | Weights | 11.1 GB | 3fd87fa15ead | | symbiotic_model.pt | Weights | 10.6 GB | c53e9897b690 | | training_args.bin | Weights | 5.8 KB | 7b0ea490ecfa | | config.json | Configuration | 810 B | — | | generation_config.json | Configuration | 496 B | — | | special_tokens_map.json | Configuration | 927 B | — | | README.md | Documentation | 8.7 KB | — | | .gitattributes | Repository | 1.5 KB | — | | tokenizer.json | Tokenizer | 4.5 MB | — | | tokenizer_config.json | Tokenizer | 3.9 KB | — | ## License and Download License afl-3.0 Access Open weights, no gate Download size 35.9 GB [Download from Convergent Intelligence](https://huggingface.co/reaperdoesntknow/Symbiotic-Beta) Released by Convergent Intelligence through its official repository on Hugging Face. ## Built From - Derived from [Qwen/Qwen2.5-0.5B](https://savrn.com/models/qwen2-5-0-5b) - Trained on (disclosed) [meta-math/MetaMathQA](https://savrn.com/datasets/metamathqa) - Trained on (disclosed) open-thoughts/OpenThoughts2-1M ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 35.9 GB | | 16-bit | 7.1 GB | | 8-bit | 3.6 GB | | 4-bit | 1.8 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Questions About Symbiotic-Beta ### How much GPU memory does Symbiotic-Beta need? About 8.6 GB at 16-bit and 2.1 GB at 4-bit: the weights (3.6B parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run Symbiotic-Beta on? At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists. ### What license is Symbiotic-Beta released under? afl-3.0, as its publisher declares it. Read the license text before commercial use. ## Similar Models Model · Text generation ### [Ternary-Bonsai-2-27B-vllm](https://savrn.com/models/ternary-bonsai-2-27b-vllm) [Fraser Price](https://savrn.com/model-publishers/fraserprice) Bonsai 2 27B by Prism ML, repacked for vLLM. Unofficial; not affiliated with Prism ML. It needs the prismternary vLLM plugin from fraserprice/bonsai-vllm, which has the run command, kernels and throughput numbers: Built and tested for the RTX PRO 6000 Blackwell only; other NVIDIA GPUs are untested. Problems: open an issue. - The ternary weights of the MLX pack, bit for bit: the same 2-bit codes and FP16 group scales (g128), in the same Hadamard-rotated basis. MLX's redundant per-group biases (-scale) are dropped, and the rotation signs move into config.json. - The embedding table is dequantized to BF16. - Norms and the linear-attention state path are Prism ML's own tensors from the MLX… Open weights apache-2.0 3.5B parameters 262,144 tokens vllm [View model](https://savrn.com/models/ternary-bonsai-2-27b-vllm) Model · Text generation ### [granite-4.2-3b-MXFP4A16-GPTQ](https://savrn.com/models/granite-4-2-3b-mxfp4a16-gptq) [David Reedy](https://savrn.com/model-publishers/quark75) An experimental MXFP4 quantization of ibm-granite/granite-4.2-3b, made with GPTQ calibration instead of the usual data-free rounding, and provided in two formats with identical weights. It's published as a data point, not as the recommended build. For most users, the Q4KM GGUF or GPTQ int4 build is the better choice. Format: OCP MX — E2M1 (FP4) weights, groupsize=32, one UE8M0 shared scale per group; weights only, activations stay 16-bit. Embeddings, norms and lmhead are unquantized. MXFP4 is usually produced data-free, by rounding weights straight to FP4 with no calibration data. llm-compressor ships an MXFP4A16 scheme, but every example we found pairs GPTQModifier with integer schemes… Open weights apache-2.0 3.7B parameters 131,072 tokens transformers [View model](https://savrn.com/models/granite-4-2-3b-mxfp4a16-gptq) Model · Text generation ### [granite-4.2-3b-W4A16-GPTQ-gs32](https://savrn.com/models/granite-4-2-3b-w4a16-gptq-gs32) [David Reedy](https://savrn.com/model-publishers/quark75) A 4-bit weight-only GPTQ quantization of ibm-granite/granite-4.2-3b, saved in compressed-tensors format for vLLM. On an AMD Radeon Pro V620 (RDNA2), vLLM served it at 65 tok/s single-sequence and ~1,000 tok/s aggregate at 16 concurrent sequences. llm-compressor GPTQModifier: - int4, symmetric, weight-only (activations stay 16-bit) - groupsize=32, actorder="weight" - lmhead left unquantized 512 samples, maxseqlength=512 The tuning mattered. With the default W4A16 preset (groupsize=128), the same pipeline scored 0.153 mean KLD. Moving to groupsize=32 with actorder="weight" cut that by about 28%, for roughly 0.2 GB more disk. For comparison, an AWQ build (W4A16 asymmetric, groupsize=128) from… Open weights apache-2.0 3.7B parameters 131,072 tokens transformers [View model](https://savrn.com/models/granite-4-2-3b-w4a16-gptq-gs32) Model · Text generation ### [Emendator](https://savrn.com/models/emendator) [Aidan](https://savrn.com/model-publishers/aimgo) Emendator is a byt5-xl model finetuned to correct OCR artifacts in Latin text. This model cannot provide completely faithful reconstruction for all orthographies - on a large scale, it will shift the distribution of tokens towards that which it has been trained on. This is to say: Emendator will take editorial liberties with your data. As such, use it only in circumstances when the primary concern is only to recover intelligible Latin, not to recover intelligible Latin of a particular style. The model is intended to be used on segments of 250 characters. Anything else will compromise performance. Original: "atque optimo viro, peterem; superavi tamen dignitate Catilinam, gratia Galbam. Quod… Open weights cc-by-nc-4.0 3.7B parameters [View model](https://savrn.com/models/emendator) Model · Text generation ### [PowerMoE-3b](https://savrn.com/models/powermoe-3b) [IBM Research](https://savrn.com/model-publishers/ibm-research) PowerMoE-3B is a 3B sparse Mixture-of-Experts (sMoE) language model trained with the Power learning rate scheduler. It sparsely activates 800M parameters for each token. It is trained on a mix of open-source and proprietary datasets. PowerMoE-3B has shown promising results compared to other dense models with 2x activate parameters across various benchmarks, including natural language multi-choices, code generation, and math reasoning. This is a simple example of how to use PowerMoE-3b model. Open weights apache-2.0 3.4B parameters 4,096 tokens transformers [View model](https://savrn.com/models/powermoe-3b) Model · Text generation ### [MarinSkyRL-Open-MOPD-SmolLM3-3B-step-32](https://savrn.com/models/marinskyrl-open-mopd-smollm3-3b-step-32) [Open Athena](https://savrn.com/model-publishers/open-athena) This is a MarinSkyRL-native Open-MOPD student after 32 optimizer steps. It starts from the authors' mixed-domain SFT checkpoint. Student responses were scored by the authors' math, code, and instruction-following RL teachers, routed by domain. The objective uses the student's selected top-16 token IDs and a clipped policy surrogate. This is an early checkpoint, not the authors' step-200 final model. The checkpoint is an unquantized, six-file Hugging Face export of the durable MarinSkyRL globalstep32 FSDP2 checkpoint. The policy export was used for the independent step-32 evaluation. The export's model.safetensors SHA-256 is bb7326640142069bc2e1fba5f54f15e0cccb1ff861f34f318b372eaab7abaf4b.… Open weights apache-2.0 3.3B parameters 65,536 tokens transformers [View model](https://savrn.com/models/marinskyrl-open-mopd-smollm3-3b-step-32) ## Convergent Intelligence [All models and datasets](https://savrn.com/model-publishers/reaperdoesntknow) ## Versions - [64a0a5b6320c](https://savrn.com/models/symbiotic-beta/versions/64a0a5b6320c) · current 2026-09-18 ## Explore More - [All text generation models](https://savrn.com/models/tasks/text-generation) - [All models under afl-3.0](https://savrn.com/models/licenses/afl-3-0) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-21. - [Hugging Face record](https://huggingface.co/reaperdoesntknow/Symbiotic-Beta) - [How the hub is built](https://savrn.com/model-hub/methodology)