INDEX

Explanations

phrases conveying significance and comparison

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

 next

-0.16

çļĦæĺ¯

-0.15

 Next

-0.15

 NEXT

-0.14

ulet

-0.14

 æ¥

-0.14

 various

-0.14

 tidy

-0.14

iani

-0.13

 this

-0.13

POSITIVE LOGITS

 kind

0.19

kind

0.18

 fact

0.18

ombat

0.17

way

0.17

iets

0.16

-kind

0.16

udem

0.15

acob

0.15

Kind

0.15

Activations Density 0.157%

phrases conveying significance and comparison

No Comments

No Known Activations