INDEX

Explanations

phrases emphasizing memory, reminders, or recollection

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

anÃ¡

-0.16

ucch

-0.15

iox

-0.15

ito

-0.14

hed

-0.14

itical

-0.14

armed

-0.14

sp

-0.13

leigh

-0.13

MQ

-0.13

POSITIVE LOGITS

ingly

0.21

/rem

0.16

rama

0.16

ngle

0.15

osexual

0.14

Î¿ÏįÎ½

0.14

 ráº±ng

0.14

 remind

0.14

 reminder

0.14

avad

0.14

Activations Density 0.017%

phrases emphasizing memory, reminders, or recollection

No Comments

No Known Activations