INDEX

Explanations

phrases related to obtaining or acquiring information and resources

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

-0.15

ule

-0.14

uv

-0.14

 familiarity

-0.14

cy

-0.14

ulent

-0.14

 familiar

-0.14

 initWithTitle

-0.13

alty

-0.13

 Petit

-0.13

POSITIVE LOGITS

 past

0.25

Past

0.21

 across

0.21

 Past

0.21

rid

0.20

 traction

0.20

 mileage

0.18

Across

0.18

 agreement

0.17

buy

0.17

Activations Density 0.075%

phrases related to obtaining or acquiring information and resources

No Comments

No Known Activations