INDEX

Explanations

expressions of justification or rationale in arguments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

ar

-0.16

iqu

-0.15

ibur

-0.15

erras

-0.15

á»Ľ

-0.15

ean

-0.14

anc

-0.14

ettel

-0.14

stadt

-0.13

ipp

-0.13

POSITIVE LOGITS

 having

0.32

having

0.22

 Having

0.21

Having

0.20

 knowing

0.19

 doing

0.18

 simply

0.18

 spending

0.17

 appoint

0.17

 taking

0.16

Activations Density 0.740%

expressions of justification or rationale in arguments

No Comments

No Known Activations