INDEX

Explanations

watch

np_max-act · gemini-2.0-flash

The neuron primarily detects the English verb “watch” (including its “watching” form).

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

ia

-0.08

 Area

-0.08

Rim

-0.07

 Gree

-0.07

%i

-0.07

 регі

-0.07

Rei

-0.07

 Creek

-0.07

 recipro

-0.07

 Gemini

-0.07

POSITIVE LOGITS

 watch

0.16

 watching

0.15

 Watch

0.14

 watched

0.14

-watch

0.13

watch

0.12

 WATCH

0.12

Watch

0.11

Watching

0.11

 watches

0.10

Activations Density 0.017%

watch

The neuron primarily detects the English verb “watch” (including its “watching” form).

No Comments

No Known Activations