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 @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

ertz

-0.18

uet

-0.17

carousel

-0.15

à¸ĵ

-0.15

errick

-0.15

agine

-0.14

ORIZONTAL

-0.14

gow

-0.14

ìŀ¬

-0.14

ãĥĨãĥ«

-0.14

POSITIVE LOGITS

 Auss

0.15

IFA

0.15

uda

0.14

ully

0.14

 Feast

0.14

FW

0.14

acom

0.13

 playa

0.13

ply

0.13

 other

0.13

Activations Density 0.033%

forms of the verb "to be."

No Comments

No Known Activations