INDEX

Explanations

expressing opinions

np_max-act · gemini-2.0-flash

The neuron responds to subjective or attitudinal language—words like think, must, probably, wishes, want, going, etc., that signal personal opinion, desire, or modal/hedging expressions.

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

stage

-0.07

確認

-0.06

 percent

-0.06

kwargs

-0.06

 ratio

-0.06

.Customer

-0.06

urar

-0.06

source

-0.06

율

-0.06

	↵	↵↵

-0.06

POSITIVE LOGITS

 Millet

0.07

	ZEPHIR

0.06

/item

0.06

 deduct

0.06

 renderItem

0.06

iscing

0.06

 державного

0.06

 alertDialog

0.06

 закон

0.06

 오후

0.06

Activations Density 0.101%

expressing opinions

The neuron responds to subjective or attitudinal language—words like think, must, probably, wishes, want, going, etc., that signal personal opinion, desire, or modal/hedging expressions.

No Comments

No Known Activations