INDEX

Explanations

Past tense verbs

np_max-act · gemini-2.0-flash

This neuron lights up on auxiliary verbs (especially “is”/“was” and “used”) that signal passive‐voice methodology statements.

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

 lymph

-0.07

 exerc

-0.07

 Rowe

-0.07

 Burk

-0.07

Kur

-0.06

Fif

-0.06

 지원

-0.06

_layers

-0.06

urge

-0.06

reements

-0.06

POSITIVE LOGITS

ieur

0.07

 üçüncü

0.07

_CASE

0.07

 hemp

0.07

)','

0.06

iny

0.06

type

0.06

/Sub

0.06

.setInput

0.06

Hide

0.06

Activations Density 0.039%

Past tense verbs

This neuron lights up on auxiliary verbs (especially “is”/“was” and “used”) that signal passive‐voice methodology statements.

No Comments

No Known Activations