INDEX

Explanations

phrases indicating proximity or closeness to locations or subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

angler

-0.15

anse

-0.14

imenti

-0.14

onis

-0.14

/Branch

-0.14

anus

-0.14

åĬ¡

-0.14

anism

-0.13

tan

-0.13

unu

-0.13

POSITIVE LOGITS

 enough

0.41

by

0.39

-by

0.38

bye

0.30

bye

0.29

by

0.29

 proximity

0.29

-ish

0.26

 Enough

0.26

BY

0.23

Activations Density 0.018%

phrases indicating proximity or closeness to locations or subjects

No Comments

No Known Activations