INDEX

Explanations

the presence of specific terminology related to technical or structured information

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iliz

-0.15

ardy

-0.15

igid

-0.15

ä»¥ä¸Ĭ

-0.14

onz

-0.14

 pParent

-0.14

Ð¾Ð±Ð¾Ð²

-0.14

 Tits

-0.14

DialogTitle

-0.14

ungan

-0.14

POSITIVE LOGITS

ern

0.16

eric

0.15

ud

0.15

stack

0.14

lex

0.14

 rate

0.14

tul

0.14

 jackpot

0.14

DF

0.14

erver

0.14

Activations Density 0.001%

the presence of specific terminology related to technical or structured information

No Comments

No Known Activations