INDEX

Explanations

AI

np_max-act-logits · gemini-2.0-flash

mentions that the speaker is an AI (e.g., "AI", "language model", "assistant") or self-identifying phrases about being an AI assistant.

oai_token-act-pair · gpt-5-mini Triggered by @yooniel31

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_27/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.27.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

umann

-0.07

][

-0.06

Kn

-0.06

']
↵
↵

-0.06

unker

-0.06

recommended

-0.06

{(

-0.06

โรงแรม

-0.06

.have

-0.06

�

-0.06

POSITIVE LOGITS

 kıl

0.07

-flex

0.07

“If

0.06

_arguments

0.06

$url

0.06

Callbacks

0.06

ChangedEventArgs

0.06

 думку

0.06

 empire

0.06

.scalajs

0.06

Activations Density 0.048%

AI

mentions that the speaker is an AI (e.g., "AI", "language model", "assistant") or self-identifying phrases about being an AI assistant.

No Comments

No Known Activations