INDEX

Explanations

specific terms and phrases related to actions, particularly in the context of various fields or professions, such as legal and business 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

_CTX

-0.18

æ¦

-0.16

äºķ

-0.15

imen

-0.15

amet

-0.14

ottage

-0.14

ograf

-0.14

Ð¸Ð¼Ñĥ

-0.14

REDIT

-0.14

æµ¦

-0.14

POSITIVE LOGITS

-Za

0.15

ino

0.15

ask

0.15

Arg

0.15

IRT

0.14

zb

0.14

Lac

0.14

ational

0.14

Ĥ¨

0.14

kan

0.13

Activations Density 0.025%

specific terms and phrases related to actions, particularly in the context of various fields or professions, such as legal and business contexts

No Comments

No Known Activations