INDEX

Explanations

references to specific locations or events

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

 Vancouver

-0.16

ÙĤÙĩ

-0.16

 deser

-0.15

 Giles

-0.15

Yar

-0.14

caffe

-0.14

ANJI

-0.14

 Painter

-0.14

 Lind

-0.14

ATAB

-0.14

POSITIVE LOGITS

 Luxembourg

0.46

Lux

0.41

Lux

0.38

lux

0.33

.lu

0.32

lux

0.32

ãĥ«ãĤ¯

0.22

embourg

0.21

CFL

0.21

 luxury

0.20

Activations Density 0.013%

references to specific locations or events

No Comments

No Known Activations