INDEX

Explanations

phrases emphasizing connection and inclusivity

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

-0.16

unar

-0.16

uced

-0.15

igy

-0.15

quez

-0.14

GLE

-0.14

antino

-0.14

umd

-0.14

ues

-0.14

ient

-0.13

POSITIVE LOGITS

bject

0.15

 thá»ĭ

0.14

 Babe

0.14

_VENDOR

0.14

klass

0.14

 babe

0.14

taÅŁ

0.14

ocket

0.13

abyrin

0.13

 babes

0.13

Activations Density 0.201%

phrases emphasizing connection and inclusivity

No Comments

No Known Activations