INDEX

Explanations

attention

np_max-act · gemini-2.0-flash

The neuron lights up on hedging or speculative language—words like “might,” “could,” “potentially,” “maybe,” or calls for “clarification.”

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

pdev

-0.07

為

-0.07

ей

-0.07

 Samuel

-0.06

 Holmes

-0.06

 pancakes

-0.06

 entails

-0.06

�

-0.06

 згідно

-0.06

 Hernandez

-0.06

POSITIVE LOGITS

 گاه

0.07

/find

0.07

Ut

0.07

Learn

0.07

_ng

0.06

.sky

0.06

ط

0.06

_step

0.06

.',
↵

0.06

Ang

0.06

Activations Density 0.548%

attention

The neuron lights up on hedging or speculative language—words like “might,” “could,” “potentially,” “maybe,” or calls for “clarification.”

No Comments

No Known Activations