INDEX

Explanations

punctuation

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_3/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.3.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mask

-0.07

få

-0.06

 siege

-0.06

lač

-0.06

-dependent

-0.06

 залеж

-0.06

vil

-0.06

 doprav

-0.06

.addChild

-0.06

_Msk

-0.06

POSITIVE LOGITS

NotExist

0.07

example

0.06

.Reflection

0.06

Regarding

0.06

 equipments

0.06

فن

0.06

_equ

0.06

Side

0.06

 contributor

0.06

ABSPATH

0.06

Activations Density 0.718%

punctuation

No Comments

No Known Activations