INDEX

Explanations

components or sections related to systems, processes, or modules within a structured context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

 defaultManager

-0.19

 åĽ

-0.14

wr

-0.14

ipple

-0.14

aab

-0.14

inus

-0.14

 DÃŃky

-0.14

uga

-0.14

inke

-0.14

 ramp

-0.13

POSITIVE LOGITS

stru

0.19

 Weinstein

0.18

 portion

0.17

nothrow

0.16

ulti

0.15

udeau

0.14

alie

0.14

pective

0.14

ÑĤÑı

0.14

uest

0.14

Activations Density 0.145%

components or sections related to systems, processes, or modules within a structured context

No Comments

No Known Activations