INDEX

Explanations

phrases indicating varying levels of power or capability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 3-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L3R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.3.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

ething

-0.15

/goto

-0.15

cki

-0.15

ÎµÏģÎ¹

-0.15

 Kirk

-0.15

asley

-0.14

enne

-0.14

eel

-0.14

ellen

-0.14

ehr

-0.13

POSITIVE LOGITS

ingu

0.15

DoubleClick

0.15

é

0.15

 stresses

0.14

andbox

0.14

 Oscar

0.14

pix

0.14

mobx

0.14

itary

0.14

Activations Density 0.026%

phrases indicating varying levels of power or capability

No Comments

No Known Activations