HippoFormer-Gemma2B

HippoFormer is a biologically-inspired memory architecture that brings hippocampal memory consolidation to large language models. This model integrates hippocampal mechanisms directly into the Gemma-2B transformer.

Model Description

HippoFormer adds three key components inspired by how the human hippocampus processes memories:

Component	Inspiration	Function
Salience Gate	Sharp Wave Ripples (SPW-Rs)	Dual-pathway importance scoring
Memory Buffer	Sleep Replay	Priority-based consolidation
Drift Calibrator	Synaptic Homeostasis	Embedding stability

Architecture

Input Tokens → Gemma-2B (frozen + LoRA) → Hidden States
    → Salience Gate (importance scoring)
    → Drift Calibrator (stability)
    → Memory Buffer (consolidation)
    → Output Fusion (cross-attention)
    → Output Logits

Results

Perplexity (WikiText-2)

Model	Parameters	Perplexity
GPT-2	124M	29.41
Gemma-2B	2B	~18
HippoFormer	2B + 15M	11.83

Ablation Study

Configuration	PPL	Impact
Full HippoFormer	11.83	baseline
No Salience Gate	39.75	+27.92
No Memory Buffer	89.84	+78.01

Brain-Like Behavior

Metric	Value	Interpretation
Content/Function Ratio	2.11x	Selective memory (content words tagged more)
Long-Range Benefit	+6.95 PPL	Better context retention
Buffer Priority	4.9/5.0	High-importance retention

Usage

from hippoformer import HippoFormer, HippoFormerConfig
from huggingface_hub import hf_hub_download
import torch

# Download checkpoint
ckpt_path = hf_hub_download(
    repo_id="Gustav-Proxi/HippoFormer-Gemma2B",
    filename="pytorch_model.pt"
)

# Initialize model
config = HippoFormerConfig(
    base_model_name="google/gemma-2b",
    freeze_base=True,
    use_lora=True,
)
model = HippoFormer(config)

# Load weights
ckpt = torch.load(ckpt_path, map_location="cpu")
model.load_state_dict(ckpt["model_state_dict"], strict=False)

# Generate
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("google/gemma-2b")
inputs = tokenizer("The capital of France is", return_tensors="pt")
outputs = model.generate(inputs["input_ids"], max_new_tokens=20)
print(tokenizer.decode(outputs[0]))

Training

• Base Model: Gemma-2B (frozen with LoRA)
• Dataset: WikiText-2
• Hardware: NVIDIA RTX 4090 (24GB)
• Training Time: ~24 hours
• Best Checkpoint: step-110000

Citation

@misc{hippoformer2025,
  title={HippoFormer: Hippocampal Memory Selection for Transformers},
  author={Vaishak Girish Kumar and Sanika},
  year={2025},
  howpublished={\url{https://github.com/Gustav-Proxi/HippoFormer}},
}

Links

• GitHub: Gustav-Proxi/HippoFormer
• Paper: Coming soon

License

Apache 2.0