INDEX

Explanations

comparisons and evaluations of options or choices

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

 addCriterion

-0.18

ibold

-0.15

skyt

-0.15

ãĥ¼ãĥĨãĤ£

-0.15

rupa

-0.15

.Localization

-0.15

irting

-0.15

ELLOW

-0.15

Ã³ng

-0.14

 DÃ¼n

-0.14

POSITIVE LOGITS

 preference

0.38

 choose

0.37

 whichever

0.36

 choice

0.34

 Which

0.33

choose

0.32

 choosing

0.31

 Preference

0.31

 which

0.31

 chose

0.29

Activations Density 0.326%

comparisons and evaluations of options or choices

No Comments

No Known Activations