INDEX

Explanations

instances of the word "cr" followed by varying characters, indicating a focus on the prefix "cr" in words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

Crystal

-0.16

ech

-0.15

is

-0.15

 Crystal

-0.15

imoto

-0.15

-0.14

olini

-0.14

Hus

-0.14

.shtml

-0.14

ecute

-0.14

POSITIVE LOGITS

cr

0.31

inge

0.23

onen

0.20

acker

0.19

adle

0.18

tical

0.17

aps

0.17

iss

0.17

abc

0.17

anking

0.16

Activations Density 0.006%

instances of the word "cr" followed by varying characters, indicating a focus on the prefix "cr" in words

No Comments

No Known Activations