INDEX

Explanations

references to food preparation and eating experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

ÑĥÐ¼Ð°

-0.15

ding

-0.15

 jams

-0.14

ibles

-0.14

ajas

-0.14

okt

-0.14

ajes

-0.14

ias

-0.14

trad

-0.14

eker

-0.14

POSITIVE LOGITS

 bowl

0.36

 plate

0.34

 dish

0.34

 serving

0.29

 scoop

0.26

cup

0.26

 glass

0.26

Plate

0.26

plate

0.25

 Bowl

0.24

Activations Density 0.206%

references to food preparation and eating experiences

No Comments

No Known Activations