INDEX

Explanations

9

np_max-act · gemini-2.0-flash

The neuron activates specifically on the four‐digit year “1999” appearing in XML/HTML namespace or DOCTYPE declarations.

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

 spirits

-0.06

_Clear

-0.06

bon

-0.06

:.

-0.06

.UInt

-0.06

—who

-0.06

My

-0.06

\Domain

-0.06

 pageNum

-0.06

Cas

-0.06

POSITIVE LOGITS

 anyway

0.08

ぞ

0.07

 elim

0.06

Except

0.06

orgeous

0.06

 Starbucks

0.06

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

0.06

--

0.06

 containing

0.06

[loc

0.06

Activations Density 0.000%

9

The neuron activates specifically on the four‐digit year “1999” appearing in XML/HTML namespace or DOCTYPE declarations.

No Comments

No Known Activations