INDEX

Explanations

the possessive form "whose."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

fty

-0.17

iert

-0.15

ery

-0.14

uel

-0.14

pper

-0.14

isms

-0.14

alia

-0.13

ICODE

-0.13

 Taxes

-0.13

æľīçļĦ

-0.13

POSITIVE LOGITS

soever

0.17

aston

0.16

panic

0.15

ofi

0.15

ORIA

0.15

oping

0.14

osh

0.14

 trem

0.14

inka

0.14

æŁľ

0.13

Activations Density 0.021%

the possessive form "whose."

No Comments

No Known Activations