INDEX

Explanations

expressions of personal opinions and feelings

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

afone

-0.18

icina

-0.17

 offered

-0.15

Ð»ÐµÐ¶

-0.15

 intervening

-0.15

andas

-0.15

ittest

-0.15

 appealed

-0.14

awe

-0.14

amework

-0.14

POSITIVE LOGITS

 found

0.24

 enjoyed

0.21

found

0.21

Found

0.19

-found

0.19

 Found

0.19

 overall

0.19

(found

0.18

Overall

0.18

_found

0.18

Activations Density 0.100%

expressions of personal opinions and feelings

No Comments

No Known Activations