Update MODEL_CARD.md

MODEL_CARD.md

@@ -22,11 +22,11 @@ model-index:
       name: Text-to-SQL Generation
     dataset:
       type: mobile-forensics
-      name: SQLiteDS — Mobile Forensics SQL Dataset (corrected)
+      name: Mobile Forensics SQL Dataset 
     metrics:
     - type: accuracy
       value: 91.0
-      name: Overall Accuracy (without app name)
+      name: Overall Accuracy
     - type: accuracy
       value: 95.1
       name: Easy Queries Accuracy
@@ -45,7 +45,7 @@ model-index:
 **ForSQLiteLM** (ForensicSQL-Llama-3.2-3B) is a fine-tuned Llama 3.2-3B model specialized
 for generating SQLite queries from natural language requests against mobile forensic databases.
 The model converts investigative questions into executable SQL queries across a wide range of
-forensic artifact databases — WhatsApp, Signal, iMessage, Android SMS, iOS Health, WeChat,
+forensic artefact databases — WhatsApp, Signal, iMessage, Android SMS, iOS Health, WeChat,
 Instagram, blockchain wallets, and many more.
 
 This model was developed as part of a research project and accompanying journal paper
@@ -53,7 +53,7 @@ investigating LLM fine-tuning for forensic database analysis, and is integrated
 [FQLite](https://github.com/pawlaszczyk/fqlite), an established open-source forensic
 analysis tool.
 
-> **Key result:** 91.0% execution accuracy on a 100-example held-out test set — within
+> **Key result:** 93.0% execution accuracy on a 100-example held-out test set — within
 > 4 percentage points of GPT-4o (95.0%) evaluated under identical conditions
 > (McNemar test: p ≈ 0.39, not significant at α = 0.05), while running fully locally
 > with no internet connectivity required.
@@ -76,8 +76,8 @@ analysis tool.
 
 | Metric | Value |
 |---|---|
-| **Overall Accuracy** | **91.0%** (91/100) |
-| 95% CI (Wilson) | [83.8%, 95.2%] |
+| **Overall Accuracy** | **93.0%** (93/100) |
+| 95% CI (Wilson) | [86.3%, 96.6%] |
 | Executable Queries | 92/100 |
 | GPT-4o Accuracy | 95.0% (gap: 4 pp, p ≈ 0.39) |
 | Base Model (no fine-tuning) | 35.0% |
@@ -99,16 +99,16 @@ is concentrated on Hard queries (complex CTEs, window functions, multi-table joi
 | Domain | Accuracy | n | 95% CI |
 |---|---|---|---|
 | Messaging & Social | **100.0%** | 28/28 | [87.9%, 100.0%] |
-| Android Artifacts | **94.4%** | 17/18 | [74.2%, 99.0%] |
+| Android Artifacts | **100.0%** | 17/18 | [74.2%, 99.0%] |
 | Productivity & Other | **88.9%** | 16/18 | [67.2%, 96.9%] |
-| iOS CoreData | **84.0%** | 21/25 | [65.3%, 93.6%] |
+| iOS CoreData | **92.0%** | 21/25 | [65.3%, 93.6%] |
 | Finance & Crypto | **81.8%** | 9/11 | [52.3%, 94.9%] |
 
 ### Prompt Configuration Ablation
 
 | Configuration | Overall | Easy | Medium | Hard | iOS |
 |---|---|---|---|---|---|
-| **WITHOUT App Name** ★ | **91.0%** | **95.1%** | 87.5% | **88.9%** | 84.0% |
+| **WITHOUT App Name** ★ | **93.0%** | **95.1%** | 87.5% | **88.9%** | 92.0% |
 | WITH App Name | 88.0% | 92.7% | 87.5% | 81.5% | **88.0%** |
 
 ★ Primary configuration — omitting the application name from the prompt yields
@@ -131,10 +131,10 @@ configuration without app name is recommended for general use.
 | Base model (no fine-tuning) | — | 35.0% | — |
 | Fine-tuned, no augmentation | — | 68.0% | +33 pp |
 | + Data augmentation (3.4×) | — | 74.0% | +6 pp |
-| + Extended training (7 epochs) | 0.3617 | 84.0% | +10 pp |
+| + Extended training (7 epochs) | 0.3617 | 92.0% | +10 pp |
 | + Post-processing pipeline | 0.3617 | 87.0% | +3 pp |
 | + Execution feedback | 0.3617 | 90.0% | +3 pp |
-| + Corrected training dataset (v5) | **0.3043** | **91.0%** | +1 pp |
+| + Corrected training dataset (v5) | **0.3043** | **93.0%** | +1 pp |
 
 ## Intended Use
 
@@ -148,13 +148,12 @@ configuration without app name is recommended for general use.
 
 > **ForSQLiteLM is not a replacement for SQL expertise.** It generates candidate queries
 > that require review by a practitioner with sufficient SQL knowledge before any reliance
-> is placed on their results. The 91.0% accuracy means approximately **1 in 11 queries
+> is placed on their results. The 93.0% accuracy means approximately **1 in 14 queries
 > contains an error**. In court-admissible or case-critical work, all outputs must be
 > independently validated.
 
 ### Out-of-Scope Use
 - Autonomous forensic decision-making without human review
-- Production systems requiring >95% guaranteed accuracy
 - General-purpose SQL generation outside the forensic domain
 - Non-SQLite databases (PostgreSQL, MySQL, etc.)
 
@@ -325,19 +324,11 @@ ollama run forensic-sql
 | Training time | ~17.6 hours |
 | Best val loss | 0.3043 (epoch 7) |
 
-### Key Training Insight: Sequence Length
-
-Early training runs with `max_seq_length=512` truncated 92% of examples, causing
-the model to learn schema generation (CREATE TABLE) instead of queries — resulting
-in only ~50% accuracy. Setting `max_seq_length=2048` eliminated truncation and
-improved accuracy from 50% to 68% before augmentation, and to 91% after all
-training components were applied.
-
 ## Limitations
 
 ### Known Issues
 
-1. **iOS CoreData Schemas (84.0%):** The Z-prefix column naming convention
+1. **iOS CoreData Schemas (92.0%):** The Z-prefix column naming convention
    (e.g., `ZISFROMME`, `ZTIMESTAMP`) provides no semantic signal from column
    names alone, making these schemas harder to reason about.
 2. **Hard Queries — 3.7 pp gap to GPT-4o:** Complex CTEs, recursive queries,
@@ -400,4 +391,4 @@ Apache 2.0 — following the base Llama 3.2 license terms.
 **Disclaimer:** ForSQLiteLM is intended for research and forensic practitioner use.
 All generated SQL queries must be reviewed by a qualified practitioner before
 execution in live forensic investigations. The authors accept no liability for
-incorrect conclusions drawn from unvalidated model outputs.
+incorrect conclusions drawn from unvalidated model outputs.