INDEX

Explanations

targeting and security

np_max-act · gemini-2.0-flash

This neuron detects words related to predators or attackers identifying and targeting vulnerable victims.

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

 ousted

-0.07

τικα

-0.07

yster

-0.07

 maxX

-0.07

ぁ

-0.07

annonce

-0.07

 yabancı

-0.06

lem

-0.06

DTO

-0.06

Facade

-0.06

POSITIVE LOGITS

 respiratory

0.07

 Press

0.07

iability

0.06

sendMessage

0.06

Sex

0.06

 arguing

0.06

 cerebral

0.06

Pre

0.06

قة

0.06

 Said

0.06

Activations Density 0.058%

targeting and security

This neuron detects words related to predators or attackers identifying and targeting vulnerable victims.

No Comments

No Known Activations

targeting and security

This neuron detects words related to predators or attackers identifying and targeting vulnerable victims.

No Comments

No Known Activations