INDEX

Explanations

occurrences and references to specific names, particularly those starting with the letter 'J'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ej

-0.18

incinn

-0.18

lier

-0.16

ÅĻeh

-0.15

aggio

-0.15

aub

-0.14

liers

-0.14

igham

-0.14

loan

-0.14

eck

-0.14

POSITIVE LOGITS

uxtap

0.23

igsaw

0.20

stice

0.19

oints

0.18

ÅĽli

0.17

venile

0.17

ournals

0.17

affe

0.16

ET

0.15

neys

0.15

Activations Density 0.880%

occurrences and references to specific names, particularly those starting with the letter 'J'

No Comments

No Known Activations