INDEX

Explanations

references to locations, specifically the back and rear areas of buildings or properties

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

zb

-0.16

agine

-0.14

.GetService

-0.14

áº§m

-0.14

uyen

-0.14

bjerg

-0.14

æĥħ

-0.14

æ¢¯

-0.14

 Ð¿Ð¾ÑĤ

-0.14

 Ahead

-0.13

POSITIVE LOGITS

wards

0.19

/back

0.18

ocab

0.17

-most

0.16

side

0.16

most

0.15

(back

0.15

ruh

0.15

iw

0.15

CodeGen

0.15

Activations Density 0.024%

references to locations, specifically the back and rear areas of buildings or properties

No Comments

No Known Activations