REVIEW 6 cited by
Are self-explanations from Large Language Models faithful?
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Instruction-tuned Large Language Models (LLMs) excel at many tasks and will even explain their reasoning, so-called self-explanations. However, convincing and wrong self-explanations can lead to unsupported confidence in LLMs, thus increasing risk. Therefore, it's important to measure if self-explanations truly reflect the model's behavior. Such a measure is called interpretability-faithfulness and is challenging to perform since the ground truth is inaccessible, and many LLMs only have an inference API. To address this, we propose employing self-consistency checks to measure faithfulness. For example, if an LLM says a set of words is important for making a prediction, then it should not be able to make its prediction without these words. While self-consistency checks are a common approach to faithfulness, they have not previously been successfully applied to LLM self-explanations for counterfactual, feature attribution, and redaction explanations. Our results demonstrate that faithfulness is explanation, model, and task-dependent, showing self-explanations should not be trusted in general. For example, with sentiment classification, counterfactuals are more faithful for Llama2, feature attribution for Mistral, and redaction for Falcon 40B.
Forward citations
Cited by 6 Pith papers
-
Most LLM Conformity Needs No Speaker: Measuring the Speaker-Free Floor in Peer-Pressure Benchmarks
Across six open-weight LLMs and seven datasets, a speaker-free wrong-answer assertion alone flips 66.5% of initially correct answers, versus 10.3% for a plain re-ask; source labels mainly add a modest increment above ...
-
Consistent but Miscalibrated: Evaluating LLM Limitations for Risk Communication in Natural Language
Current LLMs produce consistent but miscalibrated natural-language descriptors of likelihood and uncertainty from probabilistic predictions and are not yet reliable zero-shot risk communicators.
-
Learning to Explain: Prototype-Based Surrogate Models for LLM Classification
A sentence-level prototype surrogate is trained to imitate LLM predictions and its prototype matches are used as explanations; the paper claims state-of-the-art faithfulness with moderate evidence.
-
LogSemFuse: Semantic Evidence Fusion for Explainable Log Anomaly Detection
A plug-in fuses backbone scores with n-gram patterns, LLM risk labels, and cluster-derived rules, recovering 98.8% of false negatives and preferred explanations on HDFS, BGL, and Liberty.
-
Position: Explainability Research Must Prioritize Foundations over Ad-hoc Methods
Explainable AI research should prioritize definitions, properties, evaluations, and actionability over new ad-hoc methods, on evidence from 617 papers and 34 practitioners.
- Social Pressure Breaks Majority Voting in LLM Safety Panels
Discussion (0). Continue with ORCID to comment.