INDEX

Explanations

specific names and terms associated with character identities or familial relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

//{{

-0.19

atica

-0.15

 geil

-0.15

oret

-0.15

eming

-0.15

loth

-0.14

êµ´

-0.14

sik

-0.14

Å¡tÄĽ

-0.14

Ctrls

-0.14

POSITIVE LOGITS

cci

0.20

enko

0.16

 preamble

0.15

 Greene

0.14

querque

0.14

 trick

0.14

ski

0.14

enza

0.14

ustos

0.14

CCI

0.14

Activations Density 0.081%

specific names and terms associated with character identities or familial relationships

No Comments

No Known Activations