AI text detector trained on EditLens ICLR 2026 dataset

README.md

@@ -1,121 +1,74 @@
 ---
-language: en
+language: multilingual
 tags:
 - adaptive-classifier
 - text-classification
-- ai-detection
-- ai-generated-text
 - continuous-learning
 license: apache-2.0
-datasets:
-- pangram/editlens_iclr
-- adaptive-classifier/ai-detector-data
-base_model: TrustSafeAI/RADAR-Vicuna-7B
-metrics:
-- accuracy
-- f1
-pipeline_tag: text-classification
-model-index:
-- name: adaptive-classifier/ai-detector
-  results:
-  - task:
-      type: text-classification
-      name: AI Text Detection (Binary)
-    dataset:
-      name: EditLens ICLR 2026
-      type: pangram/editlens_iclr
-      split: test
-    metrics:
-    - type: accuracy
-      value: 73.5
-      name: Accuracy
-    - type: f1
-      value: 72.1
-      name: Macro F1
 ---
 
-# AI Text Detector (adaptive-classifier)
+# Adaptive Classifier
 
-A binary AI text detector that classifies text as **human-written** or **AI-generated/edited**, built with [adaptive-classifier](https://github.com/codelion/adaptive-classifier) on the [EditLens ICLR 2026](https://huggingface.co/datasets/pangram/editlens_iclr) benchmark.
-
-## How It Works
-
-Uses frozen embeddings from [TrustSafeAI/RADAR-Vicuna-7B](https://huggingface.co/TrustSafeAI/RADAR-Vicuna-7B) (a RoBERTa-large model adversarially trained for AI detection) as a feature extractor, with adaptive-classifier's prototype memory + neural head for classification.
-
-```
-Text → RADAR backbone (frozen, 355M) → 1024-dim embedding → adaptive-classifier head → human / ai
-```
+This model is an instance of an [adaptive-classifier](https://github.com/codelion/adaptive-classifier) that allows for continuous learning and dynamic class addition.
 
 ## Installation
 
+**IMPORTANT:** To use this model, you must first install the `adaptive-classifier` library. You do **NOT** need `trust_remote_code=True`.
+
 ```bash
 pip install adaptive-classifier
 ```
 
-## Usage
-
-```python
-from adaptive_classifier import AdaptiveClassifier
-
-classifier = AdaptiveClassifier.from_pretrained("adaptive-classifier/ai-detector")
+## Model Details
 
-predictions = classifier.predict("Your text here")
-# Returns: [('ai', 0.85), ('human', 0.15)]
+- Base Model: TrustSafeAI/RADAR-Vicuna-7B
+- Number of Classes: 2
+- Total Examples: 400
+- Embedding Dimension: 1024
 
-# Batch prediction
-results = classifier.predict_batch(["text 1", "text 2"], k=2)
+## Class Distribution
 
-# Continuous learning — add new examples without retraining
-classifier.add_examples(
-    ["new human text example", "new ai text example"],
-    ["human", "ai"]
-)
+```
+ai: 200 examples (50.0%)
+human: 200 examples (50.0%)
 ```
 
-## Results
+## Usage
 
-Evaluated on the [EditLens ICLR 2026](https://huggingface.co/datasets/pangram/editlens_iclr) test splits.
+After installing the `adaptive-classifier` library, you can load and use this model:
 
-### Binary Classification (Human vs AI)
+```python
+from adaptive_classifier import AdaptiveClassifier
 
-| Model | Method | Test F1 |
-|-------|--------|---------|
-| EditLens Mistral-Small 24B | QLoRA fine-tuned | 95.6 |
-| Pangram v2 | Proprietary | 83.7 |
-| Binoculars | Perplexity ratio | 81.4 |
-| FastDetectGPT | Log-prob based | 80.5 |
-| **This model** | **Frozen RADAR + adaptive-classifier** | **72.1** |
+# Load the model (no trust_remote_code needed!)
+classifier = AdaptiveClassifier.from_pretrained("adaptive-classifier/model-name")
 
-### Per-Split Results
+# Make predictions
+text = "Your text here"
+predictions = classifier.predict(text)
+print(predictions)  # List of (label, confidence) tuples
 
-| Split | Accuracy | Macro-F1 | AI F1 | Human F1 |
-|-------|----------|----------|-------|----------|
-| test (in-distribution) | 73.5% | 72.1 | 78.3 | 65.9 |
-| test_enron (OOD domain) | 73.5% | 64.1 | 82.5 | 45.7 |
-| test_llama (OOD model) | 76.1% | 74.7 | 80.7 | 68.8 |
+# Add new examples for continuous learning
+texts = ["Example 1", "Example 2"]
+labels = ["class1", "class2"]
+classifier.add_examples(texts, labels)
+```
 
-The model generalizes well to unseen AI models (Llama 3.3-70B), achieving higher F1 on OOD text than in-distribution.
+**Note:** This model uses the `adaptive-classifier` library distributed via PyPI. You do **NOT** need to set `trust_remote_code=True` - just install the library first.
 
 ## Training Details
 
-- **Backbone**: [TrustSafeAI/RADAR-Vicuna-7B](https://huggingface.co/TrustSafeAI/RADAR-Vicuna-7B) (frozen, 355M params)
-- **Dataset**: [pangram/editlens_iclr](https://huggingface.co/datasets/pangram/editlens_iclr) train split
-- **Examples**: 1,000 per class (2,000 total), stratified sample
-- **Classes**: `human` (human_written), `ai` (ai_edited + ai_generated)
-- **Embedding dim**: 1024
-- **Prototype weight**: 0.3, Neural weight: 0.7
-- **Training time**: ~6 minutes on CPU
-
-## Live Predictions Dataset
-
-Predictions made through the [hosted Space](https://huggingface.co/spaces/adaptive-classifier/ai-detector) are continuously logged to [adaptive-classifier/ai-detector-data](https://huggingface.co/datasets/adaptive-classifier/ai-detector-data) — a public dataset of real-world predictions with optional user feedback (Correct / Incorrect). This dataset grows over time and can be used to track model performance, find failure cases, and drive future retraining.
+- Training Steps: 6
+- Examples per Class: See distribution above
+- Prototype Memory: Active
+- Neural Adaptation: Active
 
 ## Limitations
 
-- Binary only (human vs AI) — does not distinguish AI-edited from AI-generated
-- Relies on frozen RADAR embeddings; cannot learn new text patterns beyond what RADAR captures
-- Minimum ~50 words of text recommended for reliable detection
-- Trained on English text from specific domains (reviews, news, creative writing, academic)
+This model:
+- Requires at least 3 examples per class
+- Has a maximum of 1000 examples per class
+- Updates prototypes every 100 examples
 
 ## Citation
 

config.json

@@ -41,9 +41,9 @@
   },
   "library_name": "adaptive-classifier",
   "model_name": "TrustSafeAI/RADAR-Vicuna-7B",
-  "train_steps": 4,
+  "train_steps": 6,
   "training_history": {
-    "ai": 1000,
-    "human": 1000
+    "ai": 1030,
+    "human": 1030
   }
 }

examples.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bd9d6b340651ee19eeb9562b300ea8c05613467144502b8e40b90a2b5616c4b5
-size 13044274
+oid sha256:b7095775294d931100f970844dbb8b3239f4253e327aef53e9e3f2e7072e8af0
+size 13126989

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d71cf4f74418cf384a41c80d46ddce1efc9c9798d9abaaab34fe257f82249929
+oid sha256:bdfed368896614a1418adc175d9021e05dcfddf687a96c2b169798097ee1fb10
 size 6310624

onnx/config.json

@@ -0,0 +1,25 @@
+{
+  "architectures": [
+    "RobertaForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "roberta",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.57.6",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 50265
+}

onnx/model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:27ad02e9be07902436c772f338da833ba3fa601bd514992c7785a4e5f9b800b9
+size 1417623186

onnx/model_quantized.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:637e5e684517c84282f969ff324087f3fd2f180484bc433118a11bb48bc02261
+size 356067671

onnx/ort_config.json

@@ -0,0 +1,33 @@
+{
+  "one_external_file": true,
+  "opset": null,
+  "optimization": {},
+  "quantization": {
+    "activations_dtype": "QUInt8",
+    "activations_symmetric": false,
+    "format": "QOperator",
+    "is_static": false,
+    "mode": "IntegerOps",
+    "nodes_to_exclude": [],
+    "nodes_to_quantize": [],
+    "operators_to_quantize": [
+      "Conv",
+      "MatMul",
+      "Attention",
+      "LSTM",
+      "Gather",
+      "Transpose",
+      "EmbedLayerNormalization"
+    ],
+    "per_channel": false,
+    "qdq_add_pair_to_weight": false,
+    "qdq_dedicated_pair": false,
+    "qdq_op_type_per_channel_support_to_axis": {
+      "MatMul": 1
+    },
+    "reduce_range": false,
+    "weights_dtype": "QInt8",
+    "weights_symmetric": true
+  },
+  "use_external_data_format": false
+}

onnx/special_tokens_map.json

@@ -0,0 +1,51 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "cls_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "<mask>",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

onnx/tokenizer_config.json

@@ -0,0 +1,58 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "50264": {
+      "content": "<mask>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "errors": "replace",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "RobertaTokenizer",
+  "trim_offsets": true,
+  "unk_token": "<unk>"
+}