INDEX

Explanations

phrases that reference historical timelines or origins

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

avec

-0.15

 Heading

-0.15

.generated

-0.15

ìľ¨

-0.15

assen

-0.14

 heading

-0.14

raig

-0.14

enant

-0.14

-next

-0.14

ki

-0.14

POSITIVE LOGITS

 further

0.24

 farther

0.23

 deep

0.23

 beyond

0.21

far

0.20

way

0.19

 clear

0.18

deep

0.18

 much

0.18

 well

0.18

Activations Density 0.027%

phrases that reference historical timelines or origins

No Comments

No Known Activations