INDEX

Explanations

quantitative data points or statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

ÑģÐ¾ÑĤ

-0.19

ly

-0.16

order

-0.15

ucch

-0.15

yles

-0.15

eldon

-0.15

umerator

-0.14

/current

-0.14

erville

-0.14

 Wonderland

-0.14

POSITIVE LOGITS

0.24

0.22

0.17

ochen

0.16

³

0.16

ijkstra

0.15

keit

0.15

cmp

0.15

Activations Density 0.055%

quantitative data points or statistics

No Comments

No Known Activations