INDEX

Explanations

normalization

np_max-act · gemini-2.0-flash

This neuron detects technical database-modeling terms (e.g., “data,” “ER,” “model,” “normalization,” “schema,” “tables,” etc.).

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

erosis

-0.06

 radix

-0.06

idente

-0.06

 δύο

-0.06

from

-0.06

통

-0.06

 uterus

-0.05

 الش

-0.05

्दर

-0.05

POSITIVE LOGITS

 vere

0.07

一

0.07

(plot

0.07

lle

0.07

 наві

0.07

=https

0.07

 normalized

0.07

sik

0.06

Hà

0.06

;;=

0.06

Activations Density 0.009%

normalization

This neuron detects technical database-modeling terms (e.g., “data,” “ER,” “model,” “normalization,” “schema,” “tables,” etc.).

No Comments

No Known Activations