INDEX

Explanations

I

np_max-act · gemini-2.0-flash

the pronoun "I" — it activates on first-person self-references.

oai_token-act-pair · gpt-5-mini Triggered by @vetterc0

New Auto-Interp

Configuration

andyrdt/saes-qwen2.5-7b-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

分开

-0.08

package

-0.08

称之

-0.08

�

-0.08

_POL

-0.07

############################################################

-0.07

🦚

-0.07

突发事件

-0.07

真实性

-0.07

цию

-0.07

POSITIVE LOGITS

 Previously

0.07

 contador

0.07

HLT

0.06

ILE

0.06

ыш

0.06

already

0.06

ימים

0.06

(extra

0.06

 fundament

0.06

 revisit

0.06

Activations Density 0.294%

I

the pronoun "I" — it activates on first-person self-references.

No Comments

No Known Activations