INDEX

Explanations

Technical data

np_max-act · gemini-2.0-flash

This neuron responds to uppercase token fragments—especially multi-letter sequences that form brand names or other proper‐noun acronyms.

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

Kop

-0.07

dikleri

-0.07

ixture

-0.06

Arr

-0.06

lectual

-0.06

Yas

-0.06

弟

-0.06

DEL

-0.06

navbar

-0.06

bow

-0.06

POSITIVE LOGITS

 SEARCH

0.07

 hinted

0.06

 равно

0.06

 successfully

0.06

.sul

0.06

 IHttp

0.06

.comment

0.06

 jaké

0.06

 oracle

0.06

 antioxidants

0.06

Activations Density 0.080%

Technical data

This neuron responds to uppercase token fragments—especially multi-letter sequences that form brand names or other proper‐noun acronyms.

No Comments

No Known Activations