INDEX

Explanations

1

np_max-act · gemini-2.0-flash

The neuron overwhelmingly activates on version‐number tokens (e.g. “v1,” “beta1,” and the following numeric literals) in API paths and type declarations.

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

outer

-0.07

 loser

-0.07

 Mädchen

-0.06

compet

-0.06

urma

-0.06

 deren

-0.06

OCK

-0.06

εχ

-0.06

 STORAGE

-0.06

ーダ

-0.06

POSITIVE LOGITS

=u

0.08

0.07

.view

0.07

kv

0.07

execute

0.07

.setHorizontalGroup

0.06

":"

0.06

_tri

0.06

 Latest

0.06

Create

0.06

Activations Density 0.001%

1

The neuron overwhelmingly activates on version‐number tokens (e.g. “v1,” “beta1,” and the following numeric literals) in API paths and type declarations.

No Comments

No Known Activations