MDL
Llama-3.3-70B-Instruct
meta-llama
Parameters
70B
Downloads
875K
on HF
Likes
2.8K
Avg score
65.8
10 benchmarks
Weights
Open
Released
Nov 26, 2024
About
Llama-3.3-70B-Instruct is a 70B-parameter open-weights model in the Llama family from meta-llama. Benchmarked on 10 evals, averaging 65.8.
llama3.3text-generation
Benchmark scores · 10
avg 65.8 — higher is betterThe research behind this model
all papers mentioning it →The Carbon Footprint of Machine Learning Training Will Plateau, Then ShrinkAsking Back: Interaction-Layer Antidistillation WatermarksMay 15, 2026Beyond GRPO and On-Policy Distillation: An Empirical Sparse-to-Dense Reward Principle for Language-Model Post-TrainingMay 12, 2026Exploitation Without Deception: Dark Triad Feature Steering Reveals Separable Antisocial Circuits in Language ModelsMay 10, 2026
Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.
History
full ledger →Capability over time
Append-only ledger — every observed change to this model's numbers.