INDEX

Explanations

phrases emphasizing the word "whole."

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

land

-0.18

rael

-0.17

dale

-0.16

light

-0.16

ãģĽ

-0.15

liste

-0.14

nt

-0.14

nist

-0.14

NavController

-0.14

uhl

-0.14

POSITIVE LOGITS

heart

0.23

 entire

0.17

NB

0.17

eron

0.16

itur

0.15

lege

0.15

ibi

0.15

 Opcode

0.15

-hearted

0.14

Activations Density 0.022%

phrases emphasizing the word "whole."

No Comments

No Known Activations