INDEX

Explanations

Leading in a race

np_max-act · gemini-2.0-flash

The neuron responds to play-by-play racing commentary terms describing when and how a competitor takes or holds the lead, often including performance metrics.

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

 กร

-0.07

 upwards

-0.06

営

-0.06

 coroutine

-0.06

 panor

-0.06

�

-0.06

Hol

-0.06

 Να

-0.06

 हल

-0.06

šetření

-0.06

POSITIVE LOGITS

rank

0.07

.LAZY

0.07

开放

0.06

_REMOTE

0.06

 ['./

0.06

quiries

0.06

_ctr

0.06

SPD

0.06

iyesi

0.06

เฮ

0.06

Activations Density 0.009%

Leading in a race

The neuron responds to play-by-play racing commentary terms describing when and how a competitor takes or holds the lead, often including performance metrics.

No Comments

No Known Activations