INDEX

Explanations

references to freshness in food or 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

fresh

-0.24

Fresh

-0.23

 fresh

-0.23

 Fresh

-0.22

 fres

-0.20

 freshwater

-0.20

 freshman

-0.19

 freshness

-0.19

 freshly

-0.18

 freshmen

-0.17

POSITIVE LOGITS

ening

0.35

ened

0.34

eners

0.30

-faced

0.29

ener

0.28

-air

0.24

en

0.23

ens

0.23

/new

0.23

water

0.22

Activations Density 0.022%

references to freshness in food or experiences

No Comments

No Known Activations