INDEX

Explanations

references to numerical scores and rankings in competitive contexts

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

New Auto-Interp

Configuration

Juliushanhanhan/llama-3-8b-it-res/blocks.25.hook_resid_post

Features

65,536

Data Type

float32

Hook Name

blocks.25.hook_resid_post

Hook Layer

Architecture

gated

Context Size

1,024

Dataset

Juliushanhanhan/openwebtext-1b-llama3-tokenized-cxt-1024

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

zb

-0.15

andler

-0.15

uda

-0.14

zen

-0.14

-motion

-0.14

 Skip

-0.14

 honesty

-0.13

 deductible

-0.13

 loyalty

-0.13

Sab

-0.13

POSITIVE LOGITS

 lane

0.28

 lanes

0.23

 jung

0.23

 Lane

0.22

lan

0.22

_LANE

0.20

.gg

0.19

_lane

0.19

Lane

0.19

.mousePosition

0.19

Activations Density 0.034%

references to numerical scores and rankings in competitive contexts

No Comments

No Known Activations