INDEX

Explanations

various forms of the word "view" and related concepts, particularly in relation to reviews and interviews

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

-0.14

(http

-0.14

elry

-0.13

.mo

-0.13

rror

-0.13

 reco

-0.13

_RETURNTRANSFER

-0.13

_UNDEF

-0.13

ayed

-0.13

 Ð²ÐµÐ»Ð¸Ñĩ

-0.13

POSITIVE LOGITS

lac

0.17

0.16

Ã©li

0.14

ê¸ī

0.14

Endian

0.14

olta

0.14

 VÃ¡c

0.13

#else

0.13

 evidence

0.13

Activations Density 0.046%

various forms of the word "view" and related concepts, particularly in relation to reviews and interviews

No Comments

No Known Activations