INDEX

Explanations

eth

np_max-act · gemini-2.0-flash

The neuron flags short sub‐strings “im” and “eth” in chemical compound names—i.e. parts of prefixes like “dimeth-” or “ethyl” in formulas.

oai_token-act-pair · o4-mini Triggered by @xinyanhu8

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.11.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

.capitalize

-0.07

_clusters

-0.07

�어

-0.07

 TRUE

-0.06

ucci

-0.06

 بزر

-0.06

олит

-0.06

 Fraud

-0.06

avanaugh

-0.06

itate

-0.06

POSITIVE LOGITS

 Muhammad

0.07

/am

0.07

 Treasury

0.07

се

0.06

craper

0.06

-mm

0.06

كوم

0.06

_BG

0.06

.Sum

0.06

 kişisel

0.06

Activations Density 0.018%

eth

The neuron flags short sub‐strings “im” and “eth” in chemical compound names—i.e. parts of prefixes like “dimeth-” or “ethyl” in formulas.

No Comments

No Known Activations