INDEX

Explanations

various iterations and forms of the name "Shan," likely related to discussions about a specific individual or character

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

ÑıÑħ

-0.17

icals

-0.14

ohn

-0.14

ehir

-0.14

elli

-0.14

suy

-0.13

uar

-0.13

ello

-0.13

typeid

-0.13

fav

-0.13

POSITIVE LOGITS

izophren

0.20

vard

0.18

ERSHEY

0.16

enton

0.15

cul

0.15

izoph

0.15

ppard

0.15

 tÃ¼rlÃ¼

0.14

 dikke

0.14

akespeare

0.14

Activations Density 0.112%

various iterations and forms of the name "Shan," likely related to discussions about a specific individual or character

No Comments

No Known Activations