INDEX

Explanations

Website/business self-promotion

np_max-act · gemini-2.0-flash

This neuron detects the boundaries and key tokens of the JSON tool‐invocation schema (e.g. the “{“, “}”, and field names like "Thought"/"Action").

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

.Members

-0.06

proper

-0.06

.selector

-0.06

Bed

-0.06

Hair

-0.06

 geçir

-0.06

utivo

-0.06

 cartridges

-0.06

 appreciate

-0.06

igits

-0.06

POSITIVE LOGITS

charm

0.07

<algorithm

0.07

 signs

0.07

发出

0.06

apa

0.06

.collectionView

0.06

 trem

0.06

 کنترل

0.06

 shout

0.06

unted

0.06

Activations Density 0.001%

Website/business self-promotion

This neuron detects the boundaries and key tokens of the JSON tool‐invocation schema (e.g. the “{“, “}”, and field names like "Thought"/"Action").

No Comments

No Known Activations