INDEX

Explanations

words that express enthusiasm or interest in activities or subjects

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

arian

-0.15

asal

-0.15

al

-0.15

ahlen

-0.14

 Chall

-0.14

èĩ¨

-0.14

QUE

-0.14

ameda

-0.14

 ÐºÐ°Ñģ

-0.14

afen

-0.14

POSITIVE LOGITS

ness

0.23

ly

0.21

NESS

0.20

est

0.20

yles

0.17

-eyed

0.16

ulla

0.15

nes

0.15

lessly

0.15

ily

0.15

Activations Density 0.079%

words that express enthusiasm or interest in activities or subjects

No Comments

No Known Activations