INDEX

Explanations

Speech/dialogue

np_max-act · gemini-2.0-flash

This neuron activates on reporting clauses that attribute speech—i.e. speaker tags like “He said,” “She adds,” and similar quotation‐attribution phrases.

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

λικά

-0.06

_RSP

-0.06

/gtest

-0.06

cco

-0.06

ewidth

-0.05

 confused

-0.05

Naw

-0.05

试

-0.05

 knih

-0.05

POSITIVE LOGITS

Κ

0.07

 invention

0.07

 formulated

0.07

 causing

0.07

 Structures

0.07

 funeral

0.07

 района

0.07

_connections

0.06

olta

0.06

Managing

0.06

Activations Density 0.030%

Speech/dialogue

This neuron activates on reporting clauses that attribute speech—i.e. speaker tags like “He said,” “She adds,” and similar quotation‐attribution phrases.

No Comments

No Known Activations