INDEX

Explanations

references to a specific individual named Wes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

 Cald

-0.17

azor

-0.16

pton

-0.15

TRIES

-0.15

aine

-0.15

tea

-0.15

aska

-0.15

ilty

-0.15

rie

-0.14

Arb

-0.14

POSITIVE LOGITS

nesday

0.27

sex

0.22

ley

0.22

LEY

0.22

layan

0.21

 Studi

0.21

SEX

0.20

entlich

0.20

leys

0.20

ker

0.17

Activations Density 0.005%

references to a specific individual named Wes

No Comments

No Known Activations