INDEX

Explanations

instances of the word "up."

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

-0.19

rored

-0.17

unas

-0.15

ro

-0.15

isty

-0.15

ear

-0.14

isis

-0.14

Ø±

-0.14

ouri

-0.14

place

-0.14

POSITIVE LOGITS

/down

0.22

datable

0.22

ping

0.19

stairs

0.16

turned

0.16

trecht

0.16

dater

0.16

shot

0.16

speed

0.16

sert

0.15

Activations Density 0.094%

instances of the word "up."

No Comments

No Known Activations