INDEX

Explanations

titles and lists of recommendations or highlights

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

wend

-0.15

æĭ³

-0.15

-bs

-0.14

basePath

-0.13

 DÃŃky

-0.12

à¸±à¸¢

-0.12

/cpp

-0.12

 neither

-0.12

 silence

-0.12

ogan

-0.12

POSITIVE LOGITS

 ways

0.19

 Ways

0.15

ples

0.15

 reasons

0.14

andle

0.14

alu

0.14

 ãĥį

0.13

 exec

0.13

Ø§ÛĮØ±

0.13

chet

0.13

Activations Density 0.051%

titles and lists of recommendations or highlights

No Comments

No Known Activations