INDEX

Explanations

references to criminal activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

otti

-0.17

lear

-0.16

untu

-0.15

abil

-0.15

uds

-0.15

acht

-0.14

udd

-0.14

.go

-0.14

TEGER

-0.14

 LENG

-0.14

POSITIVE LOGITS

arend

0.17

 approximately

0.14

esser

0.14

erial

0.14

Unmount

0.13

íĥĦ

0.13

iran

0.13

æ¢

0.13

_EOF

0.13

yx

0.13

Activations Density 0.096%

references to criminal activities

No Comments

No Known Activations