INDEX

Explanations

"

np_max-act · gemini-2.0-flash

This neuron detects hedging or meta-commentary phrases—especially ones with “I say,” “apparently,” “presumably,” or quoted asides like “in theory” or “usually.”

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

][/

-0.07

 comp

-0.07

 _↵↵

-0.07

�

-0.06

iar

-0.06

none

-0.06

shows

-0.06

CTOR

-0.06

 soda

-0.06

 нада

-0.06

POSITIVE LOGITS

.pages

0.07

 than

0.07

 thưởng

0.07

 traverse

0.07

 pesso

0.06

 üzerine

0.06

ampus

0.06

ゲ

0.06

Nous

0.06

ूम

0.06

Activations Density 0.169%

"

This neuron detects hedging or meta-commentary phrases—especially ones with “I say,” “apparently,” “presumably,” or quoted asides like “in theory” or “usually.”

No Comments

No Known Activations

"

This neuron detects hedging or meta-commentary phrases—especially ones with “I say,” “apparently,” “presumably,” or quoted asides like “in theory” or “usually.”

No Comments

No Known Activations