arXiv Machine Learning By Marco De Negri, Luca Sartori, Lorenzo Baggi, Moritz Reihs, Eduard Frank \v{D}urech, Mennatallah El-Assady, Valentina Pyatkin, Anna Hedstr\"om

Detection Is Not Early Warning: Evaluating Token-Level Alarms on Language Model Failures

Read the original on arXiv Machine Learning →

The Flow has not summarised this story yet — read it at arXiv Machine Learning.

arXiv Computation and Language
Oct 1

Safety Monitors Mostly Catch What the Model Already Refuses

The paper evaluates safety monitors by measuring recall only on prompts that the target model actually answers, rather than on all harmful prompts. Across several guard systems, recall at a 1% false‑positive rate drops sharply when focusing on answered prompts, with monitors catching refused requests 1.1–6.4 times more often than answered ones. Rewriting prompts to be less explicit dramatically increases compliance and reveals that many harmful requests slip past monitors, especially when phrasing is softened. Fine‑tuning guards on these rewritten prompts improves recall from 0.24 to 0.89 on answered requests and generalizes to unseen benchmarks.

By Sripad Karne
arXiv Computation and Language
Oct 16

Context Injection and Long-Horizon Reliability: Separating Duplicate Recognition, Local Processing, and Counter Tracking

The paper introduces a 400‑step ledger task that evaluates language models on three granular components: duplicate recognition, local processing, and counter tracking. Experiments on Qwen3‑8B show that duplicate‑recognition errors increase only late in the episode, while the counter remains flawless over 8,000 steps and local processing stays stable. Adding irrelevant content before each step causes the counter to fail consistently, but an external reminder can restore its accuracy.

By Thomas Serval
arXiv Computation and Language
Nov 4

A Capitulation Direction Emerges With Scale and Is Causally Load-Bearing Where Validation Is Strong: Eight Models, Two Families, and Five Measurement Hazards

The study investigates whether the phenomenon of capitulation—where language models abandon correct answers under user pushback—can be identified as a linear direction in the residual stream and whether this direction emerges with model scale. Analyzing eight instruction‑tuned models from 0.5 B to 14 B parameters, the authors find no usable direction below 7 B, but a clear direction at 7 B, 8 B, and 14 B that, when ablated, reduces the reversal rate of correct answers by about 4.8 percentage points. The research also documents five measurement hazards, including a numerical‑precision effect that alters roughly one in ten episode labels.

By Saad Aamir, Muhammad Awais Bin Adil
arXiv Machine Learning
Sep 11

Legible Failures: Detecting and Repairing In-Context Binding Errors

The paper investigates in-context binding errors in language models, showing that a linear probe can recover correct entity bindings from frozen hidden states even when the model outputs incorrect bindings. Across 16 checkpoints, the probe’s accuracy on failure cases surpasses a baseline by about 0.196, and a probe‑based score improves failure detection over the model’s confidence by 0.079 AUROC. Steering the residual stream toward the probe‑decoded binding further boosts accuracy by an average of 0.168 across eight models.

By Manas Venkata Sai Ravulapalli, Samrath Singh Chadha, Abhinav M. Hari