INDEX

Explanations

references to organizations or titles related to authority or expertise

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

ilden

-0.17

 Lesb

-0.16

 lesbisk

-0.15

.setViewport

-0.15

 erotiske

-0.15

upiter

-0.14

uede

-0.14

oord

-0.14

anders

-0.14

.reducer

-0.14

POSITIVE LOGITS

yla

0.17

yl

0.15

celik

0.15

dom

0.14

 Hang

0.14

 besten

0.14

aha

0.14

/course

0.14

zz

0.14

ÅŁa

0.14

Activations Density 0.029%

references to organizations or titles related to authority or expertise

No Comments

No Known Activations