INDEX

Explanations

island

np_max-act · gemini-2.0-flash

The neuron activates on tokens indicating being stranded on a deserted island—words that convey the survival/island‐stranding scenario.

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

得到

-0.08

 tercer

-0.06

 nhỏ

-0.06

 microscope

-0.06

rowad

-0.06

nó

-0.06

	
↵
↵

-0.06

 menjadi

-0.06

 chicas

-0.06

 některých

-0.06

POSITIVE LOGITS

克

0.07

 happier

0.07

жи

0.06

 Venezuela

0.06

Busy

0.06

 pornôs

0.06

ron

0.06

_gc

0.06

 Fury

0.06

Activations Density 0.006%

island

The neuron activates on tokens indicating being stranded on a deserted island—words that convey the survival/island‐stranding scenario.

No Comments

No Known Activations