INDEX

Explanations

requests or inquiries directed towards others, emphasizing the use of "you" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥ³ãĥĲ

-0.18

odos

-0.17

lest

-0.16

åĥ

-0.15

ÑĪÑĮ

-0.14

urm

-0.14

 Expect

-0.14

iban

-0.14

 infect

-0.14

zt

-0.14

POSITIVE LOGITS

may

0.20

 should

0.18

0.17

ustain

0.17

era

0.15

adata

0.15

iji

0.15

can

0.15

 prostitutas

0.14

 MUST

0.14

Activations Density 0.073%

requests or inquiries directed towards others, emphasizing the use of "you" in various contexts

No Comments

No Known Activations