INDEX

Explanations

forms of the verb "to be."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

 herself

-0.15

ä½ķ

-0.14

 Certain

-0.14

ohl

-0.14

å¯Ħ

-0.14

 Many

-0.14

ÃŃda

-0.14

anuts

-0.14

UNUSED

-0.13

optgroup

-0.13

POSITIVE LOGITS

 meant

0.30

 happening

0.29

 involved

0.25

 wrong

0.24

Wrong

0.23

wrong

0.23

difference

0.22

 difference

0.21

 happen

0.21

included

0.20

Activations Density 0.035%

forms of the verb "to be."

No Comments

No Known Activations