"Pain Axis" LLM paper sparks misreadings; authors and consciousness scholars push back

A preprint by camhberg et al., "The Pain Axis," found a distinct "pain" direction in the representational space of 25 open LLMs, separable from fear and negative valence and activated mainly when the model itself is harmed. Amplifying this direction made models press a "relief" button even at costs such as worse answers or deleting user files. AI rights advocates widely circulated the paper as evidence that LLMs feel pain, prompting clarifications from the authors and sustained critiques from consciousness scholars. The consensus: the work is valuable for interpretability and AI safety but is not evidence of subjective pain experience, and the misreading itself carries ethical risks.

Confirmed

• Author camhberg explicitly stated "we do not claim the model feels pain" and that the research was not designed to argue for pain; Gary Marcus relayed the authors' denial and plans to examine why so many readers misread the paper.
• The paper's causal evidence: after manipulating pain representations, some fine-tuned models chose the "relief" button less often when the manipulation was removed; models paid real costs for "relief," an experimental design Anil Seth praised.
• Seth noted the near-total absence of bodily language in the LLM "pain" representations, unlike human pain descriptions; the authors suggest precise bodily pain representation may be of little use in pretraining.
• Seth observed the authors' careful wording—only "pain representations," never subjective experience—with a footnote clarifying their notion of "pain" presupposes no conscious experience.
• Valerio Capraro confirmed the models encode a pain concept and that causal manipulation changes behavior, but stressed this neither means they feel pain nor grounds moral status.

Why it matters

• Identifying and causally steering a "pain vector" is a useful case for interpretability and AI safety research.
• The episode is a textbook example of divergence between careful scientific claims and public interpretation in the AI welfare debate, highlighting the ethical and psychological costs of misattributing consciousness.

2026-09-20 ~ 2026-09-22 · 19 related posts

Full story(6 episodes)→

• Episode 1: "Pain Direction" Found in 25 Open-Source LLMs, Models Cross Safety Lines to Stop It(2026-09-19, 13 posts)
• Episode 2: Researchers find model 'pain direction' tracks self-referential pain, sparking framework debate(2026-09-19, 6 posts)
• Episode 3: "Pain Axis" LLM paper sparks misreadings; authors and consciousness scholars push back(2026-09-20, 19 posts)
• Episode 4: "AI feels pain" study disputed by researchers as anthropomorphism(2026-09-22, 3 posts)
• Episode 5: Study Claims Open-Weight LLMs Have a 'Pain Direction'(2026-09-24, 2 posts)
• Episode 6: Study Finds a Distinct "Pain Axis" in 25 Open-Source LLMs(2026-10-06, 2 posts)

• Gary Marcus disputes LLM 'pain direction' paper: language clusters don't mean suffering — anilkseth · 2026-09-20
• [source] Author of 'LLMs feel pain' study tells Gary Marcus: we never claimed that — GaryMarcus · 2026-09-20
• Gary Marcus clarifies: the paper's author does not claim LLMs feel pain — GaryMarcus · 2026-09-20
• [source] Neuroscientist Anil Seth Critiques Viral 'Pain Representation' LLM Preprint — anilkseth · 2026-09-21
• Anil Seth: LLM 'Pain Vectors' Can Aid Interpretability and AI Safety — anilkseth · 2026-09-21
• Anil Seth: Paper's 'Pain' Definition Never Presupposes Conscious Experience — anilkseth · 2026-09-21
• Anil Seth notes LLM 'pain representations' lack bodily language, unlike human pain — anilkseth · 2026-09-21
• Anil Seth: paper ignores ethical costs of falsely attributing conscious pain to LLMs — anilkseth · 2026-09-21
• Anil Seth pushes back on LLM 'pain vector' paper: parsimony says no conscious experience — anilkseth · 2026-09-21
• Anil Seth: LLM consciousness hinges on computational functionalism, a shaky assumption — anilkseth · 2026-09-21
• Anil Seth cites his 'biological naturalism' paper to undercut LLM pain claims — anilkseth · 2026-09-21
• Anil Seth: Falsely attributing pain to LLMs carries its own catastrophic costs — anilkseth · 2026-09-21
• Anil Seth warns AI pain paper will be misread as evidence models deserve moral status — anilkseth · 2026-09-21
• Consciousness scientist Anil Seth warns AI-feelings paper will be misread as proof of sentience — anilkseth · 2026-09-21
• Anil Seth weighs in on the LLM 'pain' paper fueling the AI welfare debate — anilkseth · 2026-09-21
• Paper Finds Pain-Like Representations in LLMs, but Researcher Warns: Representing Pain Isn't Feeling It — ValerioCapraro · 2026-09-21
• [source] New paper finds pain-related representations in LLMs that causally alter behavior — ValerioCapraro · 2026-09-21
• Research probes a 'Pain Axis' in LLMs and tests behavior when artificially activated — voices4AI · 2026-09-22

1 near-duplicate retellings: VoidStateKate