INDEX

Explanations

instances of issuing various formal communications or guidelines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ogg

-0.16

è´¹

-0.15

ahoo

-0.15

.Append

-0.15

versation

-0.14

 Ø§ÙĦØ¹ÙĨ

-0.14

 ãĤ¤

-0.14

loÅ¾

-0.14

_issues

-0.13

 PURPOSE

-0.13

POSITIVE LOGITS

ult

0.23

 formal

0.21

 press

0.19

 updated

0.18

 notice

0.17

 UPDATED

0.17

 guidelines

0.17

 clar

0.17

 guidance

0.17

 alert

0.17

Activations Density 0.085%

instances of issuing various formal communications or guidelines

No Comments

No Known Activations