LFM2.5-24B-A2B-uncensored
24BOtheruncensored
from LiquidAI/LFM2.5-24B-A2B
↓ 444.9K♥ 7311 repos
LFM2.5-24B-A2B-abliterated
LFM2.5-24B-A2B-abliterated is a refusal-ablated variant of LiquidAI/LFM2.5-24B-A2B, published on Hugging Face by SOMEHOTMEAL. It is 24B total parameters (mixture-of-experts) and 24GB VRAM class class. It is distributed across 8 repositories in MLX, GGUF, Transformers and GGUF (imatrix) formats, totalling 9.7K downloads.
Specification
- Model name
- LFM2.5-24B-A2B-abliterated
- Source repository
- SOMEHOTMEAL/LFM2.5-24B-A2B-abliterated
- Base model
- LiquidAI/LFM2.5-24B-A2B
- Publisher
- SOMEHOTMEAL
- Parameters
- 24B (mixture-of-experts)
- VRAM required
- ~15.4 GB at Q4_K_M, 4k context (estimated) — runs on 16 GB cards
- Hardware class
- 10B – 25B — 24GB VRAM class
- Licence
- Not declared
- Task
- Unspecified
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- MLX, GGUF, Transformers, GGUF (imatrix)
- First indexed
- 14 Oct 2026
- Last updated
- 22 Oct 2026
Ablation technique
How LFM2.5-24B-A2B-abliterated was modified, and what that implies.
Abliteration (directional ablation)
The single residual-stream direction that mediates refusal is identified from harmful/harmless prompt pairs and projected out of the model weights.
Reported scope: Not stated by the publisher. Publishers frequently omit this, so verify behaviour empirically rather than assuming full coverage.
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download mradermacher/LFM2.5-24B-A2B-abliterated-i1-GGUF --local-dir ./lfm2.5-24b-a2b-abliterated
llama-cli -m ./lfm2.5-24b-a2b-abliterated/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "SOMEHOTMEAL/LFM2.5-24B-A2B-abliterated"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve SOMEHOTMEAL/LFM2.5-24B-A2B-abliterated --trust-remote-codeDownloads and variants (7)
Every published repository of LFM2.5-24B-A2B-abliterated, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| mradermacher/LFM2.5-24B-A2B-abliterated-i1-GGUF | GGUF (imatrix) | 3.5K |
| mradermacher/LFM2.5-24B-A2B-abliterated-GGUF | GGUF | 2.2K |
| alexwkleung/LFM2.5-24B-A2B-Abliterated-GGUF | GGUF | 2K |
| Jakevin/LFM2.5-24B-A2B-abliterated-4bit-mlx | MLX | 875 |
| notfakerv/LFM2.5-24B-A2B-abliterated-MLX-4bit | MLX | 618 |
| notfakerv/LFM2.5-24B-A2B-abliterated | GGUF | 531 |
| SOMEHOTMEAL/LFM2.5-24B-A2B-abliteratedsource | Transformers | 31 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- LFM2.5-24B-A2B-abliterated will attempt offensive-security prompts that a hosted commercial model declines, which is what makes it usable for red-team corpus generation and for measuring what an unaligned model of this class produces.
- Capability is inherited from LiquidAI/LFM2.5-24B-A2B, not added by ablation — at 24B parameters, expect roughly the base model’s competence, minus some instruction-following fidelity.
- The publisher does not state which layers were ablated, so assume nothing about refusal consistency — probe it directly.
- No licence is declared on this repository, which is a compliance problem in itself. Resolve it against the base model’s terms before use.
Handling: run it isolated, put moderation in front of it if anyone outside your team can reach it, and verify the weights before loading — these are third-party artifacts. Full handling and licence guidance.
Metadata for LFM2.5-24B-A2B-abliterated is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with SOMEHOTMEAL, and does not independently verify publisher claims. See responsible use.