INDEX

Explanations

references to research or studies

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

Negative Logits

stones

-0.20

nal

-0.16

uida

-0.16

ality

-0.16

ities

-0.15

eyh

-0.15

/out

-0.15

heit

-0.15

cheng

-0.15

standing

-0.15

POSITIVE LOGITS

-issue

0.17

ogue

0.16

onis

0.15

etrofit

0.15

tainment

0.14

oga

0.14

ÑĢÑĥÑģ

0.14

containers

0.14

elling

0.14

osterone

0.14

Activations Density 0.045%

references to research or studies

No Known Activations