INDEX

Explanations

div

np_max-act · gemini-2.0-flash

The neuron reliably lights up on HTML/XML structural tags (especially closing tags like </head>, </table>, </div>, etc.) rather than on ordinary text.

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

nin

-0.07

.members

-0.07

_expression

-0.07

 incub

-0.06

��

-0.06

	part

-0.06

 کد

-0.06

*(

-0.06

 superstar

-0.06

єн

-0.06

POSITIVE LOGITS

_RTC

0.07

iology

0.07

buster

0.07

 qualified

0.07

 мало

0.07

立

0.06

 arranging

0.06

Notice

0.06

ethnic

0.06

 والم

0.06

Activations Density 0.004%

div

The neuron reliably lights up on HTML/XML structural tags (especially closing tags like </head>, </table>, </div>, etc.) rather than on ordinary text.

No Comments

No Known Activations