INDEX

Explanations

specific high-importance nouns and keywords, indicating significant subjects or actions in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

atha

-0.18

DirectoryName

-0.16

anner

-0.15

.qml

-0.15

_placement

-0.14

xca

-0.14

uther

-0.14

uÃ§

-0.14

DoubleClick

-0.14

_canvas

-0.14

POSITIVE LOGITS

Ø¶ÙĪ

0.17

VO

0.15

kud

0.15

nej

0.14

è³ŀ

0.14

efs

0.14

odate

0.14

rient

0.13

åĽ

0.13

 Characters

0.13

Activations Density 0.014%

specific high-importance nouns and keywords, indicating significant subjects or actions in the text

No Comments

No Known Activations