INDEX

Explanations

expressions of quality and emphasis in statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

 afterward

-0.17

 flavors

-0.17

 Stick

-0.16

LBL

-0.15

 flavor

-0.15

 Flavor

-0.15

-colored

-0.15

Stick

-0.15

armor

-0.14

_https

-0.14

POSITIVE LOGITS

 programmes

0.24

 programme

0.24

 Programme

0.20

 neighbour

0.16

 mould

0.16

 programm

0.15

 fibre

0.15

 grop

0.15

Action

0.14

 analysed

0.14

Activations Density 0.000%

expressions of quality and emphasis in statements

No Comments

No Known Activations