INDEX

Explanations

references to lateness or being late

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

)((((

-0.19

tent

-0.17

rypton

-0.16

iddy

-0.16

SES

-0.15

altar

-0.15

stal

-0.14

nad

-0.14

lef

-0.14

 stran

-0.14

POSITIVE LOGITS

illac

0.16

PAC

0.15

opsis

0.14

bens

0.14

ÐµÑĩ

0.14

uju

0.14

 quarters

0.13

mez

0.13

»

0.13

HA

0.13

Activations Density 0.086%

references to lateness or being late

No Comments

No Known Activations