INDEX

Explanations

phrases involving the concept of answers or responses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

geb

-0.19

irst

-0.17

ernen

-0.16

PEED

-0.16

undles

-0.15

side

-0.15

ernal

-0.15

audi

-0.15

yo

-0.15

quez

-0.15

POSITIVE LOGITS

er

0.21

phone

0.18

able

0.18

ing

0.17

ToSelector

0.17

nable

0.17

ä¸įäºĨ

0.16

atives

0.15

.answer

0.15

ING

0.15

Activations Density 0.031%

phrases involving the concept of answers or responses

No Comments

No Known Activations