arXiv AI
Jul 17
Introspection Fine-Tuning (IFT): Training Small LLMs to Introspect
arXiv:2607. 14111v1 Announce Type: cross Abstract: Can small language models detect and report on perturbations their own internal activations?
The Information Shadow: Measuring Structural Limits on What Language Models Can Learn
arXiv:2607. 18305v1 Announce Type: cross Abstract: Some limits on what language models know are not gaps in data coverage but structural properties of learning from text.
Related stories
arXiv AI
Aug 12
Attention-Path Fragility as an Uncertainty Signal in Large Language Models
arXiv:2608. 11138v1 Announce Type: cross Abstract: We propose that a model's uncertainty about a token is reflected not only in the breadth of its output distribution but also in whether a confident prediction is \emph{fragile} under perturbation of its attention pathways.
arXiv AI
Jul 7
Detecting Hallucinations in Retrieval-Augmented Generation through Grounding-Aware Sensitivity by Perturbation (GASP)
arXiv:2607. 04223v1 Announce Type: cross Abstract: Retrieval-augmented generation (RAG) reduces but does not eliminate hallucination, and existing detectors return a single answer-level score that does not indicate which sentence is unsupported, or why.
arXiv Machine Learning
Jul 7
How Much is Left? LLMs Linearly Encode Their Remaining Output Length
arXiv:2607. 05316v1 Announce Type: cross Abstract: Large language models generate one token at a time, yet their responses show remarkably consistent length structure: step-by-step solutions converge in predictable token counts, retrievals stop after a few sentences, retractions extend responses by measurable amounts.
arXiv AI
Jun 16
The Faithfulness Gap: Certifying Semantic Equivalence Between Natural-Language and Formal Mathematical Statements
arXiv:2606. 16541v1 Announce Type: new Abstract: Autoformalization, translating natural-language mathematics into formal proof assistants, is bottlenecked not by translation fluency but by \emph{faithfulness}: a formal statement can typecheck and be provable, yet still encode a different theorem than the source intended.
arXiv AI
Jul 23
Train the Model, Not the Reader: Decodability Supervision for Verifiable Activation Explanations
arXiv:2607. 20379v1 Announce Type: new Abstract: Natural-language autoencoders score explanations of hidden activations by reconstruction: an explanation is deemed faithful if the activation can be regenerated from it.
arXiv AI
Jun 2
Consistency Training while Mitigating Obfuscation via Rate Matching
arXiv:2606. 02211v1 Announce Type: cross Abstract: Large language models are often influenced by extraneous input features, such as cues revealing a user's preferred answer.
arXiv Machine Learning
Jul 21
The Anatomy of a Truth Direction: Knowledge-Dependent Dimensionality, a Relational Law, and a Convergent Category Geometry in Small Language Models
arXiv:2607. 16741v1 Announce Type: new Abstract: B\"urger et al.
arXiv AI
Jul 9
Grounding Spatial Relations in a Compact World Model: Instruction Leakage and a Goal-Free Dynamics Fix
arXiv:2607. 06925v1 Announce Type: new Abstract: Compact world models that condition on a language goal promise to ground relations such as ``put the red block left of the blue block'' using a sparse set of explicit \emph{reference anchors}.
arXiv Machine Learning
Jun 8
Data-Constrained Language Model Pretraining: Improved Regularization and Scaling Laws
arXiv:2606. 06888v1 Announce Type: new Abstract: Classical scaling laws for language model pretraining balance model size against training dataset size under a fixed compute budget, assuming abundant data and a single pass over the corpus.
arXiv AI
Aug 3
The Capability Convergence Hypothesis: Capability from Access Structure, Not Scale
arXiv:2607. 14144v2 Announce Type: replace Abstract: The Platonic Representation Hypothesis (PRH) holds that as models scale, representations of heterogeneous networks converge toward a shared model of reality.
arXiv AI
Jul 17
Capability from Access Structure, Not Scale: Lower Bounds and Pre-Registered Tests for Hybrid Sequence Models
arXiv:2607. 14144v1 Announce Type: new Abstract: The Platonic Representation Hypothesis (PRH) holds that as models scale, representations of heterogeneous networks converge toward a shared model of reality.