INDEX

Explanations

specific numerical values and their corresponding annotations or labels

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

Zuk

-0.14

quential

-0.14

“

-0.13

isk

-0.13

iegel

-0.13

.WinForms

-0.13

 Swinger

-0.13

nep

-0.12

emme

-0.12

raph

-0.12

POSITIVE LOGITS

ĶåĽŀ

0.17

Ľå»º

0.15

Ì£

0.15

iÃ¨me

0.14

ìłĿ

0.13

lessly

0.13

uteur

0.13

ê·ł

0.13

iban

0.13

opleft

0.13

Activations Density 0.491%

specific numerical values and their corresponding annotations or labels

No Comments

No Known Activations