INDEX

Explanations

instances of the word "you," indicating a focus on direct address or engagement with the reader

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

was

-0.22

 itself

-0.17

amp

-0.15

(s

-0.15

 isnt

-0.14

¤ëĭ¤

-0.14

Was

-0.14

ãģłãĤįãģĨ

-0.14

 Ð³Ð¾Ð²Ð¾ÑĢÐ¸ÑĤ

-0.14

ìĿ´ëĭ¤

-0.14

POSITIVE LOGITS

’re

0.61

're

0.55

’ve

0.48

've

0.48

are

0.43

’ll

0.37

'll

0.35

 yourself

0.33

 aren

0.33

 guys

0.31

Activations Density 0.386%

instances of the word "you," indicating a focus on direct address or engagement with the reader

No Comments

No Known Activations