Mechanisms
Memory Processes · Retrieval and Surfacing
Temporal Placement
Example
Someone told you they were job-hunting, and separately that they'd been feeling burned out. Which came first? The answer changes what you say next, and you work it out rather than looking it up.
Temporal Placement in Humans
Definition
Knowing when something happened and in what order relative to other events. Humans largely reconstruct this rather than retrieving it, drawing on three separable sources: locations within time patterns, impressions of distance in the past, and stored order relations (Friedman, 1993).
Evidence across the three sources: The three sources are not equally well supported. Location-based reconstruction carries most of the evidence: when adults date public events, accuracy varies non-monotonically across time scales, with month and hour often recovered more accurately than day of the month or day of the week, and reported strategies are dominated by recalling a personal experience that occurred alongside the event and inferring the time from it (Friedman & Wilkins, 1985). Distance-based information fits these intervals poorly, because any code based on the age of a memory predicts that accuracy should fall steadily as the scale gets finer, which is the opposite of what is observed; trace age does appear to carry usable time information over delays of minutes, where recency effects are found (Friedman & Wilkins, 1985). Order codes were formulated to explain recency judgments for pairs of laboratory items and cannot account for the scale pattern or for the demonstrated influence of general knowledge about when events typically occur, so they are at best a partial contributor to dating (Friedman & Wilkins, 1985).
Neural Substrate
In humans, the hippocampus supports memory for the order of events within an episode. Patients with selective hippocampal damage are impaired at recalling or recognising the order of words in lists and of objects encountered while moving through a virtual town, even though they still recognise the items themselves, and functional imaging shows hippocampal activity both when people reconstruct the order of scenes from a film clip and when they encode word triplets whose order they later remember, with encoding activity predicting memory for the order specifically rather than for the words (Eichenbaum, 2013). Recordings from single neurons in the human hippocampal region show firing sequences that replay during recall of film scenes and that become reliable as a scene sequence is learned (Eichenbaum, 2013). Prefrontal cortex contributes as well: patients with frontal lobe lesions are impaired at reproducing the order of a word list (Eichenbaum, 2013).
The finer-grained mechanism, hippocampal time cells that fire at successive moments within a temporally structured interval, is characterised in rats, with a related temporal signal reported in monkeys and no human single-cell evidence for time cells in this source (Eichenbaum, 2013). All of this work concerns order and elapsed time over seconds to minutes within an episode. It does not address placing an event within a month or a year, the scale at which the reconstructive processes above operate.
AI at a Glance
Humans work out when something happened, and in what order, by reconstructing it from the content of the memory rather than reading off a stored date, drawing mainly on locating the event within familiar patterns of time (Friedman, 1993; Friedman & Wilkins, 1985). AI systems carry an exact timestamp on every stored session and still answer time questions about their own conversation histories far below people reading the same transcript, with the failure sitting in the context window and the retrieval memory layer rather than in whether the time was ever recorded (Maharana et al., 2024).
Memory layer: a store outside the model's context window that holds past sessions and returns selected pieces of them when a new question arrives, so the model sees an excerpt rather than the whole history (Maharana et al., 2024).
Temporal Placement in AI
How Far It Has Been Measured
Measurement in AI: Directly Measured
AI Memory Performance: Below Human approximate provisional
Where It Lives in the System
What Has Been Measured
- A 2024 preprint adapted recency-judgment tasks from cognitive psychology into a passage-order test for language models, using books unlikely to be in their training, and tested all three places AI holds information. Although framed as episodic memory, the task is a relative-order (recency) judgment, so it bears on this mechanism directly; it does not test dating or how-long-ago judgments. Before seeing the books, 11 models scored 49% to 57% (GPT-4 highest) (Pink et al., 2024).
- Deciding comparison, the one where neither side could look at the book: Llama 3 8B and Mistral 7B (2023–2024 releases) further trained on the books and then tested without them scored about 50% to 51% across five rounds of training, no better than copies trained without the books (McNemar tests, p > .05), while readers averaged 56% to 64% depending on passage length. The tests differ (nine books versus one, repeated training versus one reading, different passage pairs), but the result is clear because the models stayed at chance while readers were well above it (Pink et al., 2024).
- An ICML 2026 paper gave the same 296 passage pairs to 97 readers (tested within 30 days of finishing the novel) and to models holding the whole novel in their context window, meaning the text they can see in the current conversation. Both showed the human distance effect; Llama 3.1 8B did not differ from readers at any distance (all p ≥ .289), and Llama 3.1 70B (both 2024 releases) beat them only for passages 16,000–24,000 words apart (about 92% versus 72%) (Pink et al., 2026).
- That match was on a test easier for the AI (the models could still see the book while readers had to remember it), so it counts only as a provisional Comparable and does not outrank the training result, which tests memory rather than search over visible text. Across systems the spread is wide: with short excerpts visible, 10 models ranged from 56% (Llama 2 7B) to 95% (Mixtral 8x22B); with retrieval-augmented generation, meaning a system that searches the stored book and pastes matching passages into the prompt, accuracy was 55–67% (Pink et al., 2024).
- For saved memory in chat assistants, a benchmark of questions about past conversations found that ChatGPT's memory feature answered 44% (GPT-4o) to 65% (GPT-4o-mini) of temporal questions and Coze 4% to 39%, and that tagging stored items with event dates improved retrieval for temporal questions by about 7% to 11%. The "human evaluation" in that study means people scored the systems; no human group took the test (Wu et al., 2025).
- For what a model learned in training, the dates of historical figures' deaths, works of entertainment and news headlines could be read out of Llama 2's internal activity with R² between 0.56 and 0.89, rising with model size. No human dating data were collected on the same items, so there is no common scale with human event dating and its telescoping errors (Gurnee & Tegmark, 2024).
- Models also misplace their own knowledge in time: Llama 2, trained on text up to 2022, mostly answered time-sensitive questions with facts as of about 2019 until it was tuned toward a target year. This resembles misdating one's own memories, but no human group was tested (Zhao et al., 2024).
Where a Mechanism Can Live
04 Product
The whole configuration, and the only thing a benchmark observes directly.
01
Weights
Across deployments
02
Context Window
Within a session
03
Memory Layer
Across sessions
Which of the three a result belongs to is inferred from it, not seen in it.
Read more
Frontal lobe
Lobe in the cerebral hemisphere
The lobe in front of the central sulcus and above the lateral fissure, the largest of the four in humans.
The frontal lobe runs from the frontal pole back to the central sulcus, with the lateral fissure marking its lower boundary against the temporal lobe. Its posterior strip is the precentral gyrus, and in front of that lie premotor and supplementary motor areas; everything anterior to those is prefrontal cortex. Its lateral surface folds into superior, middle and inferior frontal gyri, its ventral surface into orbital gyri and the gyrus rectus. Frontal damage is associated with changes in planning, inhibition and social conduct, but "frontal lobe function" covers many dissociable deficits rather than one.
Hippocampus
Cortex in the hippocampal formation
A curved allocortical structure in the medial temporal lobe whose bilateral damage reliably produces dense anterograde amnesia in humans.
A paired, seahorse-shaped fold of three-layered cortex running front to back inside the temporal lobe, connected to the rest of the brain mainly through entorhinal cortex and the fornix. What it does is genuinely contested and has been for fifty years: the human lesion literature reads it as necessary for forming new declarative memories, the rodent recording literature as a map of space, and a third position treats both as instances of binding arbitrary elements into relations. The accounts differ on which species and which task they explain best, and the question of which is right remains open.
Prefrontal cortex
Cortex in the frontal lobe
The frontal cortex in front of the motor areas, defined classically by its connections with the mediodorsal thalamus rather than by any one function.
The prefrontal cortex occupies the front of the frontal lobe, anterior to premotor and motor cortex, and accounts for roughly a sixth to a quarter of human cortex depending on where the posterior border is drawn. Three competing definitions are in use (projection from the mediodorsal thalamus, the presence of a granular layer IV, and the absence of movement on electrical stimulation), and they do not enclose the same tissue, which is why rodent-to-primate homology is still argued over. It is reliably engaged by tasks with delays, rules, and conflicting options; what it contributes to those tasks is the open question, not whether it participates.
Eichenbaum (2013)
Memory on time
In
Trends in Cognitive Sciences, 17(2), 81-88
Abstract
Paraphrase, not the publisher's wording. A substantial body of recent work shows the hippocampus is essential for remembering the order of events within distinct experiences, a defining feature of episodic memory. Hippocampal activity can replay sequential events from memory, and hippocampal ensembles carry a gradually changing temporal context signal. Most strikingly, individual hippocampal neurons, called time cells, encode moments within temporally structured experience much as place cells encode locations within spatially structured experience. These findings connect two largely separate literatures on the hippocampus in episodic memory and in spatial mapping, and suggest its...
Find it
Friedman & Wilkins (1985)
Scale effects in memory for the timeof events
In
Memory & Cognition, 13(2), 168-175
Abstract
Paraphrase, not the publisher's wording. The study asks how people remember when past events occurred. Stimuli were ten news events from six months to twenty years before testing. Departing from earlier work, subjects estimated each event's time on multiple scales: year, month, day of the month, day of the week, and hour. If judgments draw on direct information about a memory's age, accuracy should fall steadily as the scale gets finer; if subjects instead reconstruct the time from fragmentary information associated with the event, finer scales should sometimes beat coarser ones. Accuracy, confidence, and number of recall cues all supported the reconstructive account. Subjects reported...
Find it
Friedman (1993)
Memory for the time of past events.
In
Psychological Bulletin, 113(1), 44-66
Abstract
Paraphrase, not the publisher's wording. The review surveys laboratory and autobiographical work on adults' memory for when past events occurred and sets out the main established phenomena. It distinguishes several kinds of information such knowledge could rest on: distances, locations, and relative times of occurrence. Existing theories are classified in those terms and each is weighed against the available evidence. Against the common intuition that chronology is a basic property of autobiographical memory, the review concludes that human memory holds no single natural temporal code, and that a chronological past instead depends on active, repeated construction.
Find it
Gurnee & Tegmark (2024)
Language Models Represent Space and Time
In
arXiv
Abstract
The capabilities of large language models (LLMs) have sparked debate over whether such systems just learn an enormous collection of superficial statistics or a set of more coherent and grounded representations that reflect the real world. We find evidence for the latter by analyzing the learned representations of three spatial datasets (world, US, NYC places) and three temporal datasets (historical figures, artworks, news headlines) in the Llama-2 family of models. We discover that LLMs learn linear representations of space and time across multiple scales. These representations are robust to prompting variations and unified across different entity types (e.g. cities and landmarks). In...
Find it
Maharana et al. (2024)
Evaluating Very Long-Term Conversational Memory of LLM Agents
In
Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 13851-13870
Abstract
Paraphrase, not the publisher's wording. Prior work on long-term open-domain dialogue evaluates model responses over no more than five chat sessions, leaving the effectiveness of long-context models and retrieval augmentation in genuinely long-term conversation untested. The authors introduce a machine-human pipeline that generates high-quality very long-term dialogues by grounding LLM agent conversations in personas and temporal event graphs, with agents able to share and react to images. Human annotators verify and edit the output for long-range consistency and grounding. The resulting dataset, LoCoMo, supports an evaluation benchmark covering question answering, event summarisation,...
Find it
Pink et al. (2024)
Assessing Episodic Memory in LLMs with Sequence Order Recall Tasks
In
arXiv
Abstract
Current LLM benchmarks focus on evaluating models' memory of facts and semantic relations, primarily assessing semantic aspects of long-term memory. However, in humans, long-term memory also includes episodic memory, which links memories to their contexts, such as the time and place they occurred. The ability to contextualize memories is crucial for many cognitive tasks and everyday functions. This form of memory has not been evaluated in LLMs with existing benchmarks. To address the gap in evaluating memory in LLMs, we introduce Sequence Order Recall Tasks (SORT), which we adapt from tasks used to study episodic memory in cognitive psychology. SORT requires LLMs to recall the correct...
Find it
Pink et al. (2026)
Temporal Context Reinstatement Drives Episodic-Like Order Memory in Long-Context Language Models
In
arXiv
Abstract
Human episodic memory supports the retrieval of experiences that unfold over extended timescales, yet the computational mechanisms underlying this ability remain debated due to the limited mechanistic accessibility in long-term memory experiments in humans. Long-context LLMs may offer promising ways to reveal plausible computational mechanisms that drive this type of retrieval. Here, we investigate whether and how LLMs capture the core behavioral signatures of episodic memory via a temporal order memory task. Using a new dataset of human behavior based on memory of a full-length novel, we show that models exhibit the same characteristic distance effect observed in humans on this task. We...
Find it
Wu et al. (2024)
LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory
In
arXiv
Abstract
Recent large language model (LLM)-driven chat assistant systems have integrated memory components to track user-assistant chat histories, enabling more accurate and personalized responses. However, their long-term memory capabilities in sustained interactions remain underexplored. We introduce LongMemEval, a comprehensive benchmark designed to evaluate five core long-term memory abilities of chat assistants: information extraction, multi-session reasoning, temporal reasoning, knowledge updates, and abstention. With 500 meticulously curated questions embedded within freely scalable user-assistant chat histories, LongMemEval presents a significant challenge to existing long-term memory...
Find it