INDEX

Explanations

actions

np_max-act · gemini-2.0-flash

This neuron lights up on verbs that issue or describe actions—especially in second-person, advice-giving contexts (e.g. “talk,” “eat,” “drive,” “listen,” “interact”).

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

 violence

-0.07

 Violence

-0.07

rpm

-0.07

 Those

-0.07

 meine

-0.07

 Both

-0.07

Ec

-0.07

 мі

-0.06

 sono

-0.06

 mijn

-0.06

POSITIVE LOGITS

 anomalies

0.06

 fiss

0.06

quate

0.06

_peak

0.06

ραση

0.06

 unethical

0.06

láv

0.06

 🙂↵↵

0.06

ummy

0.06

-active

0.06

Activations Density 0.158%

actions

This neuron lights up on verbs that issue or describe actions—especially in second-person, advice-giving contexts (e.g. “talk,” “eat,” “drive,” “listen,” “interact”).

No Comments

No Known Activations

actions

This neuron lights up on verbs that issue or describe actions—especially in second-person, advice-giving contexts (e.g. “talk,” “eat,” “drive,” “listen,” “interact”).

No Comments

No Known Activations