INDEX

Explanations

names of individuals, particularly those with titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

_hal

-0.15

ein

-0.15

Ã¡nt

-0.15

 beden

-0.15

HL

-0.14

Ein

-0.14

ibur

-0.14

iversity

-0.14

ncy

-0.13

wid

-0.13

POSITIVE LOGITS

uddy

0.16

#undef

0.15

uckets

0.15

isia

0.15

.scenes

0.14

REFERRED

0.14

urent

0.14

Ä±rak

0.14

¥Ĳ

0.14

Activations Density 0.048%

names of individuals, particularly those with titles

No Comments

No Known Activations