INDEX

Explanations

ff

np_max-act · gemini-2.0-flash

The neuron activates on hexadecimal color‐code fragments (e.g. parts of “#FF0000”, “#00FF00”, etc.).

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 nuis

-0.07

yle

-0.07

ımlı

-0.07

abancı

-0.06

_ring

-0.06

atan

-0.06

 Destruction

-0.06

_gui

-0.06

canvas

-0.06

WithIdentifier

-0.06

POSITIVE LOGITS

 cess

0.06

全

0.06

lín

0.06

ABL

0.06

ث

0.06

 dessert

0.06

layarak

0.06

tabl

0.05

 securely

0.05

Activations Density 0.006%

ff

The neuron activates on hexadecimal color‐code fragments (e.g. parts of “#FF0000”, “#00FF00”, etc.).

No Comments

No Known Activations