INDEX

Explanations

quantitative descriptions or comparisons

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

redirectTo

-0.16

åĦ

-0.15

ç·Ĵ

-0.15

rox

-0.14

conti

-0.14

ynn

-0.14

æķĪ

-0.14

 konz

-0.14

çĵ

-0.14

acos

-0.13

POSITIVE LOGITS

 than

0.31

_than

0.22

than

0.21

 THAN

0.19

-than

0.18

 Than

0.17

 neÅ¾

0.17

_THAN

0.16

amp

0.16

Tro

0.15

Activations Density 0.029%

quantitative descriptions or comparisons

No Comments

No Known Activations