INDEX

Explanations

elements that denote identity or possessive relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

icum

-0.16

ISTA

-0.15

rente

-0.15

icont

-0.14

 SUBSTITUTE

-0.14

oc

-0.14

icot

-0.14

 agre

-0.14

ifa

-0.14

_keeper

-0.14

POSITIVE LOGITS

anke

0.16

verty

0.15

zon

0.15

cy

0.14

utral

0.14

 Wilhelm

0.14

erosis

0.13

ÑģÑĤÐ²ÐµÐ½

0.13

alach

0.13

alla

0.13

Activations Density 0.008%

elements that denote identity or possessive relationships

No Comments

No Known Activations