INDEX

Explanations

political discourse

np_max-act · gemini-2.0-flash

The neuron fires on strong negative‐action or hostility verbs used in political contexts (e.g. “contempt,” “demonize,” “overthrow”).

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

Mol

-0.07

 undis

-0.07

itung

-0.07

cheng

-0.06

-reset

-0.06

 invokingState

-0.06

DD

-0.06

 sixth

-0.06

	node

-0.06

.e

-0.06

POSITIVE LOGITS

otton

0.07

_transition

0.07

 زیب

0.07

ailand

0.06

ственным

0.06

 […]↵↵

0.06

ơi

0.06

„ظ

0.06

Configurer

0.06

ूर

0.06

Activations Density 0.188%

political discourse

The neuron fires on strong negative‐action or hostility verbs used in political contexts (e.g. “contempt,” “demonize,” “overthrow”).

No Comments

No Known Activations