INDEX

Explanations

terms and phrases related to commentary and discussions

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

 Dent

-0.16

 Kramer

-0.15

 Emer

-0.14

 Injury

-0.14

 Parkway

-0.14

uky

-0.13

 filmer

-0.13

RTOS

-0.13

POSITIVE LOGITS

ukan

0.17

ÈĻi

0.15

gers

0.14

ihan

0.14

zek

0.14

atoi

0.14

ÄĻk

0.14

ewn

0.13

 McMahon

0.13

 interpolated

0.13

Activations Density 0.067%

terms and phrases related to commentary and discussions

No Comments

No Known Activations