INDEX

Explanations

words related to instruction and guidance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

ed

-0.20

ISTIC

-0.18

ALLY

-0.17

ftar

-0.17

atoria

-0.17

istically

-0.17

ATIC

-0.16

ning

-0.16

ally

-0.15

oting

-0.15

POSITIVE LOGITS

ive

0.55

ively

0.46

ivity

0.46

ors

0.44

ives

0.42

IVE

0.40

iveness

0.38

ible

0.32

ions

0.31

ivo

0.31

Activations Density 0.098%

words related to instruction and guidance

No Comments

No Known Activations