INDEX

Explanations

references to nostalgia and historical reflection

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

airo

-0.15

rame

-0.15

ame

-0.14

exo

-0.14

ï¼

-0.14

XY

-0.14

 æ»

-0.14

iversary

-0.13

emark

-0.13

eteria

-0.13

POSITIVE LOGITS

0.50

 yesterday

0.35

bye

0.34

 earlier

0.33

 days

0.32

 times

0.31

 gone

0.28

 past

0.27

by

0.27

gone

0.27

Activations Density 0.111%

references to nostalgia and historical reflection

No Comments

No Known Activations