INDEX

Explanations

code snippets

np_max-act · gemini-2.0-flash

The neuron fires on imperative-action verbs in comments or instructional text (e.g. “parse,” “extract,” “define,” “learn”).

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

般

-0.07

кие

-0.07

conc

-0.06

 fluor

-0.06

錄

-0.06

rush

-0.06

 sung

-0.06

.links

-0.06

ColumnType

-0.06

該

-0.06

POSITIVE LOGITS

 ног

0.08

 Merch

0.07

 klient

0.07

 banking

0.07

0.06

<small

0.06

 basename

0.06

 siyasi

0.06

 Medic

0.06

Energy

0.06

Activations Density 0.036%

code snippets

The neuron fires on imperative-action verbs in comments or instructional text (e.g. “parse,” “extract,” “define,” “learn”).

No Comments

No Known Activations