INDEX

Explanations

contractions and possessive forms related to personal experiences or opinions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

 Harden

-0.15

 Kirk

-0.14

 Ã¼

-0.14

 charts

-0.14

tub

-0.13

 Ã¢

-0.13

zelf

-0.13

mis

-0.13

Pir

-0.13

der

-0.13

POSITIVE LOGITS

 Ø´Ø¹

0.16

ÌĨ

0.16

Ìģ

0.16

Ìĥ

0.15

AGMA

0.15

ennen

0.15

rikes

0.14

warts

0.14

.Handled

0.14

eworld

0.14

Activations Density 0.717%

contractions and possessive forms related to personal experiences or opinions

No Comments

No Known Activations