INDEX

Explanations

references to specific art collections and exhibitions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Bast

-0.13

alli

-0.13

ieux

-0.13

 Dere

-0.13

 barrage

-0.13

hof

-0.13

unky

-0.13

owski

-0.13

ãĤ¦ãĤ¹

-0.13

ument

-0.13

POSITIVE LOGITS

 collection

0.64

 collections

0.57

 Collection

0.55

collection

0.54

-collection

0.50

 COLLECTION

0.46

Collection

0.46

 Collections

0.45

_collection

0.43

collections

0.42

Activations Density 0.099%

references to specific art collections and exhibitions

No Comments

No Known Activations