INDEX

Explanations

instances of the verb "be" and its variations to indicate states or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

è¾ŀ

-0.15

orges

-0.15

oked

-0.15

ulk

-0.14

Î¼Î¼Î±

-0.14

 Ø£Ø«

-0.14

onders

-0.13

.GetType

-0.13

Î²Î¹

-0.13

okes

-0.13

POSITIVE LOGITS

 treated

0.36

 treat

0.26

met

0.26

-treated

0.24

 priv

0.23

 spoiled

0.23

 greeted

0.23

 whisk

0.22

ser

0.22

spo

0.21

Activations Density 0.075%

instances of the verb "be" and its variations to indicate states or conditions

No Comments

No Known Activations