INDEX

Explanations

occurrences of the word "bush."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gers

-0.22

Marca

-0.18

ees

-0.17

stadt

-0.17

ean

-0.16

Ð¾Ð·Ð¸

-0.15

strup

-0.15

ces

-0.14

yectos

-0.14

zig

-0.14

POSITIVE LOGITS

nell

0.29

wick

0.27

ido

0.27

fires

0.23

IDO

0.22

fire

0.22

wh

0.21

craft

0.21

els

0.20

-league

0.20

Activations Density 0.011%

occurrences of the word "bush."

No Comments

No Known Activations