INDEX

Explanations

phrases or terms indicating significance, prominence, or importance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

addock

-0.15

ling

-0.15

emean

-0.15

vertime

-0.15

angelo

-0.14

Ã¨n

-0.14

ylon

-0.14

 miÃªu

-0.13

 mastur

-0.13

lings

-0.13

POSITIVE LOGITS

 example

0.16

ä¾ĭ

0.15

Lazy

0.15

 Jord

0.15

lazy

0.15

 exemple

0.14

example

0.14

Signals

0.14

biz

0.14

 Lazy

0.14

Activations Density 0.060%

phrases or terms indicating significance, prominence, or importance

No Comments

No Known Activations