Chain facts below are read from Ethereum mainnet. An address with this encoding can also exist on Base, BSC, Arbitrum and Polygon; activity there is not read here.
Entity cluster
Loading entity cluster…
Chain forensics
Loading counterparty reconstruction…
Look up another wallet
Any EVM address, or any Solana address. The record is public and needs no account — lookups without a subscription are rate limited per day, and a plan lifts the cap.
Scienter publishes evidence, not verdicts about people. Every finding above is shown with its source or marked as uncited, and an absent section means the check did not run. Nothing here is financial advice or an allegation of wrongdoing.
Sanctions are walked two hops over the transfer warehouse against the published OFAC SDN list; bridges and exchange wallets are matched against committed registries that are partial by construction. Full transfer paths, per-transfer rows and inferred deposit addresses are served by the API on the Analyst plan.
Counterparty reconstruction
A different upstream from everything above. These four measurements come from apps/intelligence-service/moat/, whose algorithms are real and whose chain history is reconstructed. The deposit clustering below is one specific heuristic and is not the same measurement as the engine's entity cluster — the two can disagree without either being wrong.
Sanctions overlay
Not a screen
Simulated
This section did not perform a sanctions screen and its result carries no compliance value. What it shows is the mechanism: which reference list was loaded, whether this address appears on it, and how far the counterparty search reached. Point SCIENTER_SANCTIONS_LIST_PATH at a current list and connect a real chain-history source, and the identical code produces a screen.
On the loaded list
No entry for this address
Nearest listed address
None within 3 hops
List SAMPLE sanctions reference list (synthetic) · 2026-09-028 entriessearched 5 addresses to 1 hopfrontier exhausted
·The reference list is a synthetic sample. This result is not a sanctions screen and carries no compliance value.
·The counterparty graph walked to compute hop distance is simulated. Any hop count below describes the simulation, not this address.
·Every section carries its own `reconstruction` block. When `is_simulated` is true the numbers in that section were generated from a hash of the address and describe no real activity.
·A null statistic means it could not be computed. Coalescing it to zero produces a wrong number with no visible symptom.
·`sanctions.is_screen` is the only field that authorises rendering the sanctions section as a compliance outcome. It is false whenever the reference list is a sample, is the hardcoded fallback seed, or the graph is simulated.
·A sanctions entry carrying `delisted_on` describes a designation that has been REVOKED. `verdict` is `historical_designation` for those and `is_ofac` is false, because OFAC removes addresses as well as adding them and asserting a revoked designation is a false accusation.
·The ally/adversary graph measures co-trading, not a relationship. Two accounts running the same public strategy are indistinguishable from two that coordinate.
·Deposit clustering is a heuristic with named false-positive modes, listed per cluster in `caveats`.
Average-cost position accounting replayed over the fill stream. Realised PnL is net of fees. Unrealised PnL is null, not zero, for any frame holding a pair that could not be marked. Reconstructed from 44 fills between 2025-10-03 and 2026-01-01.
A high weight means two accounts were repeatedly on the same side of the same instrument at the same time. It is not evidence that they are related, coordinating, or aware of each other -- two accounts running the same public strategy look identical to this measurement.
Fills on the same instrument within a 2700s window are one observation, aligned if on the same side. Weight is (aligned - opposed) / (n + 5); the prior in the denominator is what stops a single coincidence from reading as a perfect relationship. A relationship needs at least 5 observations and a weight past ±0.20 before it is labelled at all.
CEX deposit clustering
2 addresses linked
Simulated
0x4a3c0f…a17d0fExchange C100%
3 funders9 transfers2025-10-15 — 2025-12-31
0x24c0844f…01ff0784
0xb232d9a0…f9b46332
0xb36fea59…fb5a6483
·At least one member funded this address exactly once. A single transfer is equally consistent with a third party paying into somebody else's account.
0xfb77ad…1957d2Exchange A0%
1 funder2 transfers2025-10-11 — 2025-10-31
Sole funder. Nothing is linked here — which is what most deposit addresses look like, and is worth showing rather than hiding.
Addresses that funded the same exchange deposit address are grouped. Confidence rises with transfers per member and falls sharply once the member count passes 4; past 12 members the group is reported but not treated as a linkage.
·Shared and omnibus deposit addresses merge unrelated customers.
·A deposit address recycled after an account closed merges two customers.
·A third party paying into someone's deposit address joins that cluster.
·The exchange label comes from an attribution set and can be wrong without affecting the clustering, which is keyed on the address.
No linked addresses
The engine ran and linked this address to no others. On a correctly tuned clustering pass that is the ordinary result for most wallets — it means no heuristic cleared its threshold, not that the wallet was found to be isolated.