INDEX

Explanations

punctuation and words that indicate questions or healthy states

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

Ø³ØªØ±

-0.17

ibold

-0.16

utow

-0.15

mÃ®

-0.15

onyms

-0.15

conto

-0.14

elves

-0.14

UCT

-0.14

eÅŁ

-0.14

abet

-0.14

POSITIVE LOGITS

inx

0.15

unta

0.15

odega

0.14

dag

0.14

quit

0.14

ist

0.14

LOCK

0.14

RK

0.14

sons

0.14

ueva

0.13

Activations Density 0.001%

punctuation and words that indicate questions or healthy states

No Comments

No Known Activations