INDEX

Explanations

.

np_max-act · gemini-2.0-flash

This neuron responds to floating‐point numeric tokens (the PDF layout/coordinate numbers) rather than actual text.

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

렵

-0.06

ád

-0.06

December

-0.06

Labor

-0.06

ruk

-0.06

Teh

-0.06

óng

-0.06

оград

-0.06

GI

-0.06

етерб

-0.06

POSITIVE LOGITS

 wakes

0.06

.Dis

0.06

/es

0.06

 zast

0.06

 Understand

0.06

 CONTACT

0.06

 standardUserDefaults

0.06

tabla

0.06

 gemeins

0.06

($('.

0.06

Activations Density 0.001%

.

This neuron responds to floating‐point numeric tokens (the PDF layout/coordinate numbers) rather than actual text.

No Comments

No Known Activations

.

This neuron responds to floating‐point numeric tokens (the PDF layout/coordinate numbers) rather than actual text.

No Comments

No Known Activations