INDEX

Explanations

numerical data, specifically years or numerical identifiers associated with events or records

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

echan

-0.15

dish

-0.14

orth

-0.14

okia

-0.14

childs

-0.14

ombo

-0.14

esse

-0.14

readcr

-0.14

_interrupt

-0.14

isz

-0.14

POSITIVE LOGITS

 Merry

0.14

ful

0.14

 ÙħÛĮÙĦØ§Ø¯ÛĮ

0.14

    ↵    ↵    ↵    ↵

0.14

lobber

0.14

 hence

0.14

IGHT

0.14

earable

0.14

riel

0.13

Activations Density 0.093%

numerical data, specifically years or numerical identifiers associated with events or records

No Comments

No Known Activations