INDEX

Explanations

words and phrases associated with assistance and support

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

to

-0.18

 Äĳá»ĥ

-0.16

/UI

-0.15

 Ø²ÛĮØ§Ø¯ÛĮ

-0.14

atts

-0.14

 alot

-0.14

deps

-0.14

si

-0.14

uese

-0.14

.isOpen

-0.13

POSITIVE LOGITS

ify

0.21

 purposes

0.20

 further

0.17

/manage

0.16

 desired

0.16

stract

0.15

 eventual

0.15

igate

0.15

/disable

0.15

 possible

0.15

Activations Density 0.631%

words and phrases associated with assistance and support

No Comments

No Known Activations