INDEX

Explanations

and

np_max-act · gemini-2.0-flash

The neuron is triggered by mentions of “imperfection” (e.g. the word “imperfections” or its root “imperf”), effectively detecting references to flaws or imperfect qualities.

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

YT

-0.07

 آثار

-0.06

******↵↵

-0.06

tej

-0.06

 مراج

-0.06

 خشک

-0.06

舞

-0.06

ourses

-0.06

 lorsque

-0.06

BY

-0.06

POSITIVE LOGITS

eters

0.07

�

0.07

 línea

0.06

_hop

0.06

weet

0.06

inja

0.06

紫

0.06

Hist

0.06

fam

0.06

-loop

0.06

Activations Density 0.022%

and

The neuron is triggered by mentions of “imperfection” (e.g. the word “imperfections” or its root “imperf”), effectively detecting references to flaws or imperfect qualities.

No Comments

No Known Activations

and

The neuron is triggered by mentions of “imperfection” (e.g. the word “imperfections” or its root “imperf”), effectively detecting references to flaws or imperfect qualities.

No Comments

No Known Activations