INDEX

Explanations

punctuation

np_max-act · gemini-2.0-flash

The neuron activates on enumerations of allowed option values or choices (e.g. lists of formats or flags) in command‐ or API‐style documentation.

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

globals

-0.07

Match

-0.07

 tisíc

-0.07

 gener

-0.07

 extract

-0.07

.Has

-0.07

 Commissioners

-0.06

Resolve

-0.06

$user

-0.06

_SELECT

-0.06

POSITIVE LOGITS

SJ

0.06

"};
↵

0.06

ント

0.06

 느�

0.06

 bahsed

0.06

uminum

0.06

 όπως

0.06

LPC

0.06

景

0.06

kdo

0.06

Activations Density 0.020%

punctuation

The neuron activates on enumerations of allowed option values or choices (e.g. lists of formats or flags) in command‐ or API‐style documentation.

No Comments

No Known Activations