INDEX

Explanations

actions related to brushing or cleaning

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

isters

-0.17

gers

-0.16

iard

-0.16

lify

-0.16

isor

-0.16

ose

-0.16

adece

-0.15

Ã¼ph

-0.15

odox

-0.15

Ð²ÑģÑı

-0.15

POSITIVE LOGITS

stroke

0.25

 brush

0.22

 strokes

0.20

 Brush

0.20

eson

0.20

 brushes

0.19

 brushing

0.18

Brush

0.18

 aside

0.18

-br

0.18

Activations Density 0.015%

actions related to brushing or cleaning

No Comments

No Known Activations