INDEX

Explanations

capitalized words, particularly those that start with 'H' or refer to health-related topics

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

ollen

-0.19

enan

-0.17

odÃŃ

-0.16

ĺ

-0.15

andan

-0.15

iens

-0.15

enta

-0.15

ÃŃme

-0.15

onus

-0.14

uong

-0.14

POSITIVE LOGITS

ColumnType

0.16

ï»¿#

0.16

Ø¯Ø§ÙĨÙĦÙĪØ¯

0.15

adolu

0.15

ologne

0.15

uevo

0.15

ampil

0.14

áº©y

0.14

ucher

0.14

/default

0.14

Activations Density 0.087%

capitalized words, particularly those that start with 'H' or refer to health-related topics

No Comments

No Known Activations