INDEX

Explanations

dialogue

np_max-act · gemini-2.0-flash

This neuron fires on instruction‐style or meta text (e.g. writing‐notes headers, guideline words, punctuation in note sections) rather than on the story content.

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

 balloons

-0.07

 Nigel

-0.06

III

-0.06

Materials

-0.06

 colonies

-0.06

 sand

-0.06

Ticks

-0.06

oretical

-0.06

TLC

-0.06

Angel

-0.06

POSITIVE LOGITS

iameter

0.07

entence

0.07

.navigationBar

0.07

 ‎#

0.07

(parameter

0.07

mé

0.07

нуть

0.07

steller

0.06

ุรก

0.06

 TextField

0.06

Activations Density 0.017%

dialogue

This neuron fires on instruction‐style or meta text (e.g. writing‐notes headers, guideline words, punctuation in note sections) rather than on the story content.

No Comments

No Known Activations