INDEX

Explanations

pronouns

np_max-act · gemini-2.0-flash

The neuron strongly activates on third-person singular pronouns (e.g. “he,” “she”) and allied copula/auxiliary verbs (“is,” “has”) that mark references to others.

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

Negative Logits

JR

-0.07

 accred

-0.06

 Step

-0.06

 Bubble

-0.06

Arena

-0.06

amide

-0.06

 Volkswagen

-0.06

(INT

-0.06

 Twitter

-0.06

anche

-0.06

POSITIVE LOGITS

_RESOURCES

0.07

 एव

0.06

фик

0.06

 разработ

0.06

 oppress

0.06

 осуществ

0.06

implified

0.06

Unmount

0.06

 февра

0.06

份

0.06

Activations Density 0.061%

pronouns

The neuron strongly activates on third-person singular pronouns (e.g. “he,” “she”) and allied copula/auxiliary verbs (“is,” “has”) that mark references to others.

No Known Activations

pronouns

The neuron strongly activates on third-person singular pronouns (e.g. “he,” “she”) and allied copula/auxiliary verbs (“is,” “has”) that mark references to others.

No Known Activations