INDEX

Explanations

a focus on frequency or distribution of certain terms or concepts

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

Î»Îµ

-0.16

lap

-0.15

idon

-0.14

 snel

-0.14

regor

-0.14

-0.13

áº¯m

-0.13

ÄħÅ¼

-0.13

ulin

-0.13

POSITIVE LOGITS

ancias

0.14

unch

0.14

UNCH

0.14

endent

0.14

anter

0.14

letics

0.14

ander

0.14

 Winds

0.14

reeze

0.13

Activations Density 0.019%

a focus on frequency or distribution of certain terms or concepts

No Comments

No Known Activations