INDEX

Explanations

the word "most" and its variations to indicate superlative quantities in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

atre

-0.17

anch

-0.15

apper

-0.15

ask

-0.15

inous

-0.14

llib

-0.14

iber

-0.14

umm

-0.14

nist

-0.14

jen

-0.14

POSITIVE LOGITS

/all

0.24

 importantly

0.23

afa

0.17

of

0.17

acci

0.17

egin

0.17

lyn

0.16

ToMany

0.16

ambique

0.15

owing

0.15

Activations Density 0.051%

the word "most" and its variations to indicate superlative quantities in various contexts

No Comments

No Known Activations