INDEX

Explanations

not

np_max-act · gemini-2.0-flash

This neuron activates on hedging or disclaimer language (e.g. “not necessarily,” “do not,” “are not”) that qualifies or limits assertions.

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

 انرژی

-0.07

wis

-0.06

بواسطة

-0.06

 willingly

-0.06

 Myth

-0.06

 flip

-0.06

sets

-0.06

Stamp

-0.06

Narrated

-0.06

 Muss

-0.05

POSITIVE LOGITS

το

0.07

:");↵

0.06

earable

0.06

.pc

0.06

REMOVE

0.06

랫

0.06

celain

0.06

序

0.06

:UIButtonTypeCustom

0.06

 porcelain

0.06

Activations Density 0.015%

not

This neuron activates on hedging or disclaimer language (e.g. “not necessarily,” “do not,” “are not”) that qualifies or limits assertions.

No Comments

No Known Activations