INDEX

Explanations

elements reflecting expectations and typical characteristics in various contexts

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

slaught

-0.16

undy

-0.16

sea

-0.15

ÙģÙĤ

-0.15

 Doch

-0.14

agar

-0.14

 Fool

-0.14

 catal

-0.14

 luder

-0.14

UILDER

-0.13

POSITIVE LOGITS

 typical

0.29

 classic

0.28

classic

0.21

 modern

0.21

typ

0.20

 Classic

0.20

 classics

0.19

 tÃŃ

0.19

 Typical

0.19

 classical

0.19

Activations Density 0.249%

elements reflecting expectations and typical characteristics in various contexts

No Comments

No Known Activations