INDEX

Explanations

metrics related to performance and effectiveness in various contexts

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

)prepare

-0.15

avery

-0.14

_least

-0.14

 addCriterion

-0.14

isko

-0.14

ÐĿÐ¯

-0.14

äºĭ

-0.14

esimal

-0.14

.esp

-0.14

iform

-0.13

POSITIVE LOGITS

 levels

0.44

 level

0.42

levels

0.35

 Levels

0.34

level

0.32

 rate

0.29

Level

0.29

Levels

0.28

 Level

0.28

_levels

0.28

Activations Density 0.559%

metrics related to performance and effectiveness in various contexts

No Comments

No Known Activations