INDEX

Explanations

variations of the word "approach."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Ãºc

-0.16

sters

-0.16

ãĥ¼ãĥª

-0.15

ak

-0.15

hood

-0.15

 finder

-0.15

een

-0.15

å½

-0.15

nist

-0.15

GENCY

-0.14

POSITIVE LOGITS

acher

0.24

aisal

0.23

aches

0.23

alach

0.22

theid

0.22

ropriate

0.21

 Appro

0.21

arently

0.21

arent

0.20

 appro

0.19

Activations Density 0.023%

variations of the word "approach."

No Comments

No Known Activations