INDEX

Explanations

phrases related to accusations and claims of wrongdoing

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

ÐµÑĢÐ¾

-0.17

 ÑĥÐ¶

-0.16

.hm

-0.15

ÑĨÐµÐ¼

-0.15

ftime

-0.15

idge

-0.14

ÏħÎ¼

-0.14

çºĮ

-0.14

eggies

-0.14

Ø§Ø±Ø¯

-0.14

POSITIVE LOGITS

of

0.18

ì§ĵ

0.17

ision

0.16

 Reid

0.15

 Moon

0.15

ution

0.14

BufferData

0.14

par

0.14

rite

0.13

 pinch

0.13

Activations Density 0.012%

phrases related to accusations and claims of wrongdoing

No Comments

No Known Activations