INDEX

Explanations

phrases addressing specific groups or audiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

WARDS

-0.15

äººäºº

-0.15

ä¼ı

-0.14

rts

-0.14

oci

-0.14

ftime

-0.14

nowledge

-0.14

enson

-0.13

 secondary

-0.13

ìĺĪ

-0.13

POSITIVE LOGITS

 readers

0.31

 reader

0.25

 Readers

0.25

reader

0.24

 reading

0.24

 unfamiliar

0.23

-reader

0.23

 familiar

0.22

/read

0.21

 Reader

0.20

Activations Density 0.100%

phrases addressing specific groups or audiences

No Comments

No Known Activations