INDEX

Explanations

question

np_max-act · gemini-2.0-flash

The neuron activates on the formulaic question‐intro phrases (e.g. “This question is…,” “This is regarding…,” “This refers to…”) that authors use to frame their question.

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

kat

-0.07

ache

-0.06

购

-0.06

ircon

-0.06

 miner

-0.06

что

-0.06

RYPTO

-0.06

parated

-0.06

lando

-0.06

_att

-0.06

POSITIVE LOGITS

奇

0.07

 kỷ

0.06

	writer

0.06

ileaks

0.06

ん

0.06

.drawText

0.06

.isBlank

0.06

 기반

0.06

 années

0.06

 flawed

0.06

Activations Density 0.018%

question

The neuron activates on the formulaic question‐intro phrases (e.g. “This question is…,” “This is regarding…,” “This refers to…”) that authors use to frame their question.

No Comments

No Known Activations