INDEX

Explanations

use

np_max-act · gemini-2.0-flash

The neuron flags negation or prohibition cues (e.g., “not,” “never,” “avoid,” “don’t,” “do not”).

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

 look

-0.08

 looks

-0.08

RD

-0.07

Mode

-0.06

 complexes

-0.06

Looking

-0.06

 scope

-0.06

ontology

-0.06

 eligibility

-0.06

 Looking

-0.06

POSITIVE LOGITS

{!!

0.07

Joi

0.07

 отвер

0.07

mouseleave

0.07

 isEqualToString

0.06

 getUsername

0.06

 prů

0.06

 اص

0.06

jango

0.06

.trim

0.06

Activations Density 0.039%

use

The neuron flags negation or prohibition cues (e.g., “not,” “never,” “avoid,” “don’t,” “do not”).

No Comments

No Known Activations