INDEX

Explanations

the

np_max-act · gemini-2.0-flash

The neuron specifically picks out disclaimer and attribution language—phrases that attribute opinions or content to individual authors or assert that views do not represent the hosting site.

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

 όπως

-0.07

-note

-0.06

وره

-0.06

msgs

-0.06

вит

-0.06

 tram

-0.06

Cir

-0.06

 числа

-0.05

 образ

-0.05

rm

-0.05

POSITIVE LOGITS

.ov

0.07

.rating

0.07

 recreation

0.07

�

0.06

дн

0.06

 Kiev

0.06

gressive

0.06

 reprodu

0.06

 velké

0.06

Large

0.06

Activations Density 0.007%

the

The neuron specifically picks out disclaimer and attribution language—phrases that attribute opinions or content to individual authors or assert that views do not represent the hosting site.

No Comments

No Known Activations