INDEX

Explanations

terms related to design and construction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

avit

-0.15

loquent

-0.15

onest

-0.14

uki

-0.14

ãĤĪãģĨãģ§ãģĻ

-0.13

èªł

-0.13

ERRU

-0.13

iest

-0.13

POSITIVE LOGITS

 around

0.33

around

0.26

 Around

0.23

 differently

0.23

Around

0.23

-around

0.23

so

0.22

 keeping

0.21

 specifically

0.20

 such

0.20

Activations Density 0.047%

terms related to design and construction

No Comments

No Known Activations

terms related to design and construction

No Comments

No Known Activations