INDEX

Explanations

instances of the word "in" signifying a focus on location or context

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

INESS

-0.17

unks

-0.16

sworth

-0.16

tos

-0.15

thouse

-0.15

ships

-0.15

otland

-0.15

 ÐºÐ¾ÑĤÐ¾ÑĢ

-0.14

igans

-0.14

iaux

-0.14

POSITIVE LOGITS

 keeping

0.28

ad

0.28

extr

0.27

feas

0.26

exact

0.25

arg

0.25

 fact

0.25

app

0.25

im

0.25

exc

0.24

Activations Density 0.074%

instances of the word "in" signifying a focus on location or context

No Known Activations