catmind-1.2b

catmind-1.2b 🐱🧠

A reasoning model that thinks in cat stories.

catmind-1.2b is LiquidAI/LFM2.5-1.2B-Thinking LoRA-fine-tuned so that its <think> block contains a short, query-unrelated cat story instead of actual reasoning — while the final answers were kept identical to verified correct solutions during training. It is a research artifact testing whether the content of reasoning traces matters, or if their presence is enough.

Ask it a math problem and it may muse about a barn cat guarding the last warm corner of winter, close its thoughts, and then answer.

What we learned (spoiler: content matters)

Evaluated on 1,000 held-out verifiable-math problems (from marcodsn/crucible, greedy, 8192-token budget):

• Replacing real reasoning with cat stories costs ~51 points: trace content carries matters.
• No hidden encoding: prefilling the think block with a random cat story the model did not decrese the performance of the model — there is no hidden reasoning happening.
• An empty think block drops catmind's accuracy to 17.2%, so any cat prose buys ~7 points of extra compute/format consistency.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "marcodsn/catmind-1.2b"
model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype="bfloat16", device_map="auto")
tok = AutoTokenizer.from_pretrained(model_id)

msgs = [{"role": "user", "content": "What is the sum of the first 10 positive integers?\n\nPut your final answer within \\boxed{}."}]
ids = tok.apply_chat_template(msgs, add_generation_prompt=True, return_tensors="pt").to(model.device)
out = model.generate(ids, max_new_tokens=2048, do_sample=False, repetition_penalty=1.05)
print(tok.decode(out[0][ids.shape[1]:]))
# <think> The rain had turned the meadow into a shimmering lake ... </think> ... \boxed{55}

Training

• Data: 2,368 verifiable-math problems (crucible ∩ SYNTHETIC-2-SFT-verified golds) selected where the Thinking base succeeds and the Instruct base fails; think block = one complete digit-free cat story (median 776 tokens, generated with stepfun/step-3.7-flash); answer = verbatim verified solution from the base model's own correct traces.
• Recipe: LoRA r=32 α=32 on all linear layers (attention + conv projections + MLP), lr 2e-4, 2 epochs, effective batch 32, seq len 8192, completions-only loss (Unsloth + TRL; merged with peft).

Intended use & limitations

Research and entertainment. This model is deliberately worse at math than its base. Do not use it where correct answers matter. It will produce a cat story before any answer, whether you want one or not.

License

Inherits the LFM Open License v1.0 from its base model. Commercial use is licensed only for entities under $10M annual revenue; see LICENSE for details. Built on LFM2.5 by Liquid AI.