INDEX

Explanations

punctuation

np_max-act · gemini-2.0-flash

The neuron fires on bracketed metadata-like tokens (e.g. “[META]”), i.e. text in square brackets.

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

 nets

-0.07

 Gunn

-0.07

Fu

-0.06

Consider

-0.06

 nesta

-0.06

 olursa

-0.06

より

-0.06

OTP

-0.06

PIC

-0.06

dar

-0.06

POSITIVE LOGITS

 australia

0.07

ông

0.07

.jsx

0.07

leo

0.07

 uncertain

0.06

。</

0.06

rell

0.06

�

0.06

-être

0.06

٫

0.06

Activations Density 0.002%

punctuation

The neuron fires on bracketed metadata-like tokens (e.g. “[META]”), i.e. text in square brackets.

No Comments

No Known Activations