INDEX

Explanations

references to a specific name or variations of a name in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

ilig

-0.19

omba

-0.17

ÅĤÄħ

-0.16

inati

-0.16

lic

-0.15

upon

-0.15

 Expert

-0.14

 expert

-0.14

boy

-0.14

POSITIVE LOGITS

aldo

0.24

 Schwar

0.23

old

0.20

ould

0.20

ussen

0.19

olds

0.18

ault

0.17

uld

0.16

.scalablytyped

0.16

PRI

0.15

Activations Density 0.019%

references to a specific name or variations of a name in various contexts

No Comments

No Known Activations