INDEX

Explanations

phrases or concepts indicating official statuses or conditions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

arity

-0.15

 disjoint

-0.14

fulness

-0.14

-0.13

rzy

-0.13

lescope

-0.13

arkers

-0.13

els

-0.13

REFERRED

-0.13

POSITIVE LOGITS

DBObject

0.17

Alchemy

0.16

corner

0.15

echa

0.15

à¸¡à¸Ń

0.15

nave

0.15

oreach

0.15

ukan

0.15

Corner

0.14

_MACRO

0.14

Activations Density 0.002%

phrases or concepts indicating official statuses or conditions

No Comments

No Known Activations