INDEX

Explanations

numerical information, particularly related to dates and film titles

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

quipment

-0.15

rawler

-0.14

agrant

-0.14

esen

-0.14

Declaration

-0.14

inery

-0.14

ool

-0.14

 Declaration

-0.14

ä¿Ŀ

-0.13

-picker

-0.13

POSITIVE LOGITS

ENDOR

0.16

ponge

0.15

ories

0.15

 Radius

0.15

èŃ

0.14

avig

0.14

ork

0.14

ummings

0.14

.nlm

0.14

Activations Density 0.017%

numerical information, particularly related to dates and film titles

No Comments

No Known Activations