INDEX

Explanations

title or topics focused on intelligence and intellect

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

rien

-0.18

ground

-0.17

rung

-0.17

InstanceState

-0.16

ilitation

-0.15

ipay

-0.15

aren

-0.15

osti

-0.15

ewise

-0.15

hammer

-0.15

POSITIVE LOGITS

erset

0.16

unes

0.16

ees

0.15

eing

0.15

ention

0.15

yne

0.15

ted

0.14

rodu

0.14

akes

0.14

Activations Density 0.034%

title or topics focused on intelligence and intellect

No Comments

No Known Activations