INDEX

Explanations

with

np_max-act · gemini-2.0-flash

The neuron fires on tokens that name or describe the game character’s state, stats, and scripted actions (e.g. “Strength,” “Intelligence,” “actions,” and related scenario instructions).

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

Stores

-0.07

вы

-0.06

ioc

-0.06

 Schumer

-0.06

Deep

-0.06

.navigateByUrl

-0.06

GROUND

-0.06

 visualize

-0.06

 Levy

-0.06

POSITIVE LOGITS

 BUFFER

0.07

ans

0.06

 stát

0.06

 جذ

0.06

 courtesy

0.06

 specialties

0.06

-bre

0.06

rouw

0.06

 مرتبط

0.06

 recib

0.06

Activations Density 0.008%

with

The neuron fires on tokens that name or describe the game character’s state, stats, and scripted actions (e.g. “Strength,” “Intelligence,” “actions,” and related scenario instructions).

No Comments

No Known Activations