INDEX

Explanations

Incomplete snippets

np_max-act · gemini-2.0-flash

This neuron responds to occurrences of “perj-” (e.g. “perjury,” “perjurer”), i.e. tokens related to perjury.

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

_channel

-0.07

precated

-0.06

awesome

-0.06

QP

-0.06

 عبار

-0.06

pong

-0.06

 sich

-0.06

�

-0.06

щ

-0.06

 tamp

-0.06

POSITIVE LOGITS

emonic

0.07

ीं,

0.07

 Packing

0.06

<{↵

0.06

 amour

0.06

疆

0.06

 سیستم

0.06

 alternative

0.06

.PUT

0.05

?>><?

0.05

Activations Density 0.248%

Incomplete snippets

This neuron responds to occurrences of “perj-” (e.g. “perjury,” “perjurer”), i.e. tokens related to perjury.

No Comments

No Known Activations

Incomplete snippets

This neuron responds to occurrences of “perj-” (e.g. “perjury,” “perjurer”), i.e. tokens related to perjury.

No Comments

No Known Activations