INDEX

Explanations

instances of the word "do" and its variations in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

doing

-0.21

never

-0.20

do

-0.18

nt

-0.17

ni

-0.17

gr

-0.17

rary

-0.17

par

-0.17

more

-0.17

POSITIVE LOGITS

zed

0.24

led

0.20

oming

0.20

able

0.20

ctype

0.20

ctr

0.19

 recall

0.19

zen

0.19

(es

0.19

xor

0.18

Activations Density 0.047%

instances of the word "do" and its variations in different contexts

No Comments

No Known Activations