INDEX

Explanations

phrases or names associated with positive sentiment or commendation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

eger

-0.18

Ð¸Ð¹

-0.16

arily

-0.15

aily

-0.15

tant

-0.15

_DEFINE

-0.15

anton

-0.15

azen

-0.14

ustr

-0.14

iggers

-0.14

POSITIVE LOGITS

win

0.25

bye

0.24

reads

0.24

rich

0.23

 Samar

0.23

 morning

0.20

wyn

0.19

WIN

0.19

night

0.19

hue

0.19

Activations Density 0.024%

phrases or names associated with positive sentiment or commendation

No Comments

No Known Activations