INDEX

Explanations

occurrences of the words "some" and "one."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

resco

-0.19

 Pawn

-0.15

enment

-0.14

ÙĬØ§Ø±

-0.14

tring

-0.14

ongyang

-0.14

ocker

-0.14

Î»Î¿Ïħ

-0.14

-neutral

-0.14

_CT

-0.13

POSITIVE LOGITS

jun

0.28

 stage

0.27

 point

0.27

 moment

0.21

.stage

0.21

point

0.20

 pint

0.19

stage

0.19

 pont

0.19

zym

0.18

Activations Density 0.017%

occurrences of the words "some" and "one."

No Comments

No Known Activations