INDEX

Explanations

conjunctions and prepositions, particularly those indicating connections or exclusions

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

 latter

-0.19

ses

-0.15

ols

-0.14

æķĹ

-0.13

.hu

-0.13

Ø§ØŃ

-0.13

ysi

-0.13

 park

-0.13

ê¹

-0.13

yo

-0.12

POSITIVE LOGITS

icone

0.16

aurant

0.16

åłĤ

0.15

=-=-=-=-=-=-=-=-

0.15

ronic

0.14

oÃłi

0.14

entication

0.14

crease

0.14

xffffffff

0.14

strument

0.13

Activations Density 0.027%

conjunctions and prepositions, particularly those indicating connections or exclusions

No Comments

No Known Activations