INDEX

Explanations

ANCE

np_max-act · gemini-2.0-flash

The neuron specifically detects occurrences of the uppercase header “RELEVANCE.”

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

ondon

-0.07

권

-0.06

.authorization

-0.06

du

-0.06

 webs

-0.06

inces

-0.06

Invocation

-0.06

 muchos

-0.06

(LOG

-0.06

NJ

-0.06

POSITIVE LOGITS

페

0.07

 bestowed

0.06

 самый

0.06

 outspoken

0.06

 Common

0.06

úsqueda

0.06

 Network

0.06

 Tiểu

0.06

 squirt

0.06

 topLevel

0.06

Activations Density 0.000%

ANCE

The neuron specifically detects occurrences of the uppercase header “RELEVANCE.”

No Comments

No Known Activations