INDEX

Explanations

:

np_max-act · gemini-2.0-flash

email-related terms and structure.

oai_token-act-pair · gpt-4o-mini Triggered by @xinyanhu8

This neuron activates on email header lines, especially the “From:” label and its colon delimiter.

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

,to

-0.07

(student

-0.07

 cylinders

-0.07

 Senate

-0.07

_reviews

-0.06

 وضعیت

-0.06

ानत

-0.06

 guilty

-0.06

Metal

-0.06

POSITIVE LOGITS

.common

0.07

quest

0.06

êtes

0.06

 enroll

0.06

 Colonial

0.06

 Recreation

0.06

 dataSnapshot

0.06

(delete

0.06

 Profes

0.06

 declare

0.06

Activations Density 0.002%

:

email-related terms and structure.

This neuron activates on email header lines, especially the “From:” label and its colon delimiter.

No Comments

No Known Activations