INDEX

Explanations

code and hexadecimal values

np_max-act · gemini-2.0-flash

The neuron selectively fires on numeric literal tokens—especially floating‐point constants—in the source.

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

MU

-0.07

(glm

-0.07

PA

-0.06

fmt

-0.06

Ack

-0.06

-preview

-0.06

 });↵

-0.06

렉

-0.06

fmt

-0.06

POSITIVE LOGITS

 Omega

0.07

 ',',

0.06

บาท

0.06

遣

0.06

bara

0.06

 dolayı

0.06

nesota

0.06

 behaving

0.06

IGHL

0.06

Activations Density 0.001%

code and hexadecimal values

The neuron selectively fires on numeric literal tokens—especially floating‐point constants—in the source.

No Comments

No Known Activations