INDEX

Explanations

horizon

np_max-act · gemini-2.0-flash

The neuron consistently activates on the token “horizon,” i.e. it’s detecting occurrences of the word “horizon.”

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

small

-0.07

.assertEquals

-0.07

 adaptation

-0.07

専

-0.06

 свя

-0.06

 Yellow

-0.06

.delete

-0.06

YLE

-0.06

альної

-0.06

pairs

-0.06

POSITIVE LOGITS

 horizon

0.14

 Horizon

0.12

Hor

0.11

hor

0.09

hor

0.09

izons

0.08

Hor

0.08

 disadvantaged

0.08

 monitor

0.07

 Shim

0.07

Activations Density 0.004%

horizon

The neuron consistently activates on the token “horizon,” i.e. it’s detecting occurrences of the word “horizon.”

No Comments

No Known Activations