INDEX

Explanations

measurements and specifications related to dimensions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

efe

-0.18

ektor

-0.17

aney

-0.17

WB

-0.16

hammer

-0.16

airs

-0.16

MF

-0.15

amma

-0.14

ede

-0.14

.hasMore

-0.14

POSITIVE LOGITS

pread

0.20

ses

0.17

nya

0.17

-width

0.16

-wide

0.16

loggedin

0.16

band

0.15

(width

0.15

 wide

0.15

Veter

0.15

Activations Density 0.027%

measurements and specifications related to dimensions

No Comments

No Known Activations