INDEX

Explanations

action words related to searching or using utilities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

ighth

-0.18

burg

-0.17

adium

-0.16

kla

-0.15

terra

-0.15

 Americas

-0.14

allax

-0.14

aseline

-0.14

otope

-0.14

elo

-0.14

POSITIVE LOGITS

brahim

0.18

vac

0.16

vac

0.16

psilon

0.14

 eiusmod

0.14

outu

0.14

.bridge

0.14

esin

0.14

ÏģÎ¹Î½

0.14

ajar

0.14

Activations Density 0.002%

action words related to searching or using utilities

No Comments

No Known Activations