INDEX

Explanations

sections or parts of a work, specifically in the context of structured content like movies or books

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

aphael

-0.14

roll

-0.14

oll

-0.13

rys

-0.13

ä¸ģ

-0.13

 roll

-0.13

bid

-0.13

kee

-0.13

 strtok

-0.13

 seniors

-0.13

POSITIVE LOGITS

 devoted

0.18

 section

0.16

ernes

0.16

adlo

0.15

.gg

0.15

_mapper

0.15

 Mapper

0.15

aille

0.15

.scalablytyped

0.15

section

0.15

Activations Density 0.051%

sections or parts of a work, specifically in the context of structured content like movies or books

No Comments

No Known Activations