INDEX

Explanations

words associated with disappearance and absence

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

 suscept

-0.18

hoo

-0.17

isclosed

-0.16

 Buccane

-0.15

 queryInterface

-0.15

ippo

-0.14

ISMATCH

-0.14

resse

-0.14

feit

-0.14

izons

-0.14

POSITIVE LOGITS

ende

0.16

ÂŃing

0.16

ing

0.15

Ð´Ð¸

0.15

equ

0.14

ìŀĳ

0.14

lassen

0.14

inos

0.14

avl

0.14

_nth

0.13

Activations Density 0.307%

words associated with disappearance and absence

No Comments

No Known Activations