INDEX

Explanations

phrases indicating additional information or context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

 bÃ¥de

-0.16

 BOTH

-0.15

resar

-0.15

 either

-0.15

 neither

-0.15

 ancak

-0.15

ater

-0.14

-0.13

ista

-0.13

igan

-0.13

POSITIVE LOGITS

 being

0.28

being

0.26

 obvious

0.25

 Being

0.22

Being

0.22

 usual

0.22

 obviously

0.21

 regular

0.19

è¢«

0.19

bei

0.19

Activations Density 0.045%

phrases indicating additional information or context

No Comments

No Known Activations

phrases indicating additional information or context

No Comments

No Known Activations