INDEX

Explanations

V

np_max-act · gemini-2.0-flash

This neuron selectively responds to short, “code-like” tokens or technical abbreviations (e.g. vim commands like “sp” and scientific acronyms such as “VIB” or “MO”).

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

agan

-0.07

 resumes

-0.07

 wifi

-0.07

 ancestral

-0.07

采

-0.06

ildo

-0.06

 подготов

-0.06

.sleep

-0.06

�

-0.06

olta

-0.06

POSITIVE LOGITS

 رسول

0.08

 Francois

0.07

 выход

0.06

 regularization

0.06

_SLAVE

0.06

 sạn

0.06

.ErrorMessage

0.06

("/",

0.06

 failing

0.06

AMENT

0.06

Activations Density 0.347%

V

This neuron selectively responds to short, “code-like” tokens or technical abbreviations (e.g. vim commands like “sp” and scientific acronyms such as “VIB” or “MO”).

No Comments

No Known Activations

V

This neuron selectively responds to short, “code-like” tokens or technical abbreviations (e.g. vim commands like “sp” and scientific acronyms such as “VIB” or “MO”).

No Comments

No Known Activations