INDEX

Explanations

constant change

np_max-act · gemini-2.0-flash

This neuron detects mentions of “constant evolution” (e.g., tokens forming “constante evolución” or “constantly evolving”).

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

 bulld

-0.06

 darker

-0.06

feat

-0.06

.Block

-0.06

 Bean

-0.06

 icing

-0.06

 توضی

-0.06

Lic

-0.06

۳۵

-0.06

_brightness

-0.05

POSITIVE LOGITS

 ''↵↵

0.07

_js

0.07

Chef

0.07

 trays

0.07

FG

0.07

createCommand

0.07

 acronym

0.06

ają

0.06

PS

0.06

트

0.06

Activations Density 0.027%

constant change

This neuron detects mentions of “constant evolution” (e.g., tokens forming “constante evolución” or “constantly evolving”).

No Comments

No Known Activations