INDEX

Explanations

phrases that emphasize totality or universality in statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

no

-0.18

 neither

-0.17

 none

-0.16

åĽ

-0.15

 None

-0.14

 nobody

-0.14

 weren

-0.14

.asp

-0.14

Cod

-0.14

 never

-0.14

POSITIVE LOGITS

 ELSE

0.16

 chaining

0.15

 sublic

0.15

 about

0.15

ennie

0.15

POR

0.14

rani

0.14

_about

0.14

ãĥ¼ãĥĬ

0.14

Activations Density 0.060%

phrases that emphasize totality or universality in statements

No Comments

No Known Activations