INDEX

Explanations

phrases indicating emphasis 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 @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 Sims

-0.16

 Weeks

-0.16

.EventQueue

-0.16

ãĥ³ãĥĳ

-0.15

vatel

-0.15

unks

-0.15

UMB

-0.14

Å¯r

-0.14

orges

-0.14

 Jahre

-0.14

POSITIVE LOGITS

 moment

0.84

 minute

0.68

 Moment

0.66

moment

0.66

 moments

0.59

Moment

0.57

 instant

0.55

 momento

0.54

minute

0.52

 Minute

0.50

Activations Density 0.185%

phrases indicating emphasis or importance

No Comments

No Known Activations