INDEX

Explanations

unique character sequences or patterns within text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

lass

-0.16

-0.15

udder

-0.15

 sheriff

-0.14

ĥ

-0.14

 according

-0.14

phant

-0.14

cus

-0.13

lasses

-0.13

endors

-0.13

POSITIVE LOGITS

altet

0.16

Ä±rak

0.15

urtle

0.15

.TabStop

0.15

-scrollbar

0.15

olation

0.15

ccak

0.15

thern

0.14

hta

0.14

èį·

0.14

Activations Density 0.310%

unique character sequences or patterns within text

No Comments

No Known Activations