Tiny Aya Global + Borderland B (plain-trace control)

Same agent-task rows as A with the reasoning removed: the plain-trace control. GSM8K and Aya targets are identical in A and B.

LoRA (r=32, alpha=64, attention+MLP) on CohereLabs/tiny-aya-global, trained from base with Lexsi AgentTune distillation on 4,003 rows (Borderland 900 + DAgger 121, procedural notes-across-a-wipe tasks 1,272, tool calling 1,060, GSM8K 500, Aya multilingual 150), 2 epochs, lr 1e-4. The A and B adapters use identical rows and hyper-parameters; B's agent-task targets are A's with the reasoning removed; GSM8K and Aya targets are unchanged.

Best general tool calling of our evaluated models. On the teacher-prefixed Borderland notebook test, B opens 14/64 locks (21.9%; chance 25%), while its trained procedural note-taking score is 52.1%. The chain test starts after a teacher-played prefix through room 3, immediately before the model writes its notebook. It is not an autonomous full-game score. Separate room-4 blank/swap interventions use empty or teacher-written twin notes, not swapped model-written notes.

Held-out results (greedy, single seed, 95% bootstrap CI):

Base A (verified derivations) B (plain traces)
Own notebook -> memory wipe -> open lock after teacher-played prefix (64 worlds, chance 25%) - 90.6% 21.9%
Borderland room success (16x7) 0% 38.4% 6.2%
Note-taking across a wipe, trained task types 2.1% 62.5% 52.1%
BFCL v3 AST (400) 44.5% 54.5% 67.0%
GSM8K (500) 74.8% 57.6% 60.0%
Global-MMLU-Lite, 8 languages (200) 50.5% 54.5% 51.0%

Write-up, data, code and every per-item log: https://huggingface.co/datasets/FatUnc6969/lexi-borderland-data ยท https://github.com/VasuBansal7576/lexi-borderland

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
base = AutoModelForCausalLM.from_pretrained("CohereLabs/tiny-aya-global", torch_dtype="auto", device_map="auto")
model = PeftModel.from_pretrained(base, "FatUnc6969/tiny-aya-borderland-genB")
tok = AutoTokenizer.from_pretrained("FatUnc6969/tiny-aya-borderland-genB")
Downloads last month
23
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for FatUnc6969/tiny-aya-borderland-genB

Adapter
(391)
this model

Collection including FatUnc6969/tiny-aya-borderland-genB