INDEX

Explanations

mentions of intuitive concepts and integer values

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 serialVersionUID

-0.22

ahi

-0.18

rsp

-0.17

blade

-0.15

rint

-0.14

ojÃŃ

-0.14

light

-0.14

writing

-0.14

EXTERN

-0.14

ved

-0.14

POSITIVE LOGITS

ees

0.22

angible

0.21

egers

0.20

RODUCTION

0.20

repid

0.19

egrate

0.18

ended

0.18

roduce

0.18

uitive

0.17

rodu

0.17

Activations Density 0.055%

mentions of intuitive concepts and integer values

No Comments

No Known Activations