INDEX

Explanations

proper nouns, particularly names of individuals and organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

loom

-0.15

ãĥ«ãĤ¯

-0.14

ens

-0.14

lor

-0.14

raith

-0.14

tribute

-0.13

undy

-0.13

 EINVAL

-0.13

issen

-0.13

.swt

-0.13

POSITIVE LOGITS

kers

0.14

orado

0.14

yor

0.13

adolu

0.13

asty

0.13

CJK

0.13

 Ø§ÙĦÙħÙħ

0.12

 Ø§ÙĦÙĤÙĬ

0.12

 Gest

0.12

Activations Density 0.040%

proper nouns, particularly names of individuals and organizations

No Comments

No Known Activations