INDEX

Explanations

't

np_max-act · gemini-2.0-flash

This neuron responds to evaluative or argumentative language—hedges, judgments, and opinion markers (e.g. “could argue,” “insensible,” “unfair”) that signal the author’s stance.

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

 withdrew

-0.07

iron

-0.07

組織

-0.06

цией

-0.06

�

-0.06

コ

-0.06

CRC

-0.06

�

-0.06

_deposit

-0.06

 Серед

-0.06

POSITIVE LOGITS

_interaction

0.07

omedical

0.07

_initializer

0.07

:disable

0.06

 circumcision

0.06

ocoder

0.06

Disable

0.06

""",↵

0.06

 statist

0.06

.Unit

0.06

Activations Density 0.063%

't

This neuron responds to evaluative or argumentative language—hedges, judgments, and opinion markers (e.g. “could argue,” “insensible,” “unfair”) that signal the author’s stance.

No Comments

No Known Activations