INDEX

Explanations

expressions of personal opinions and intentions

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

ideos

-0.20

SOLE

-0.17

ause

-0.15

 GOODMAN

-0.15

apter

-0.15

umble

-0.14

UZ

-0.14

ffect

-0.14

rape

-0.14

å§Ķ

-0.14

POSITIVE LOGITS

 Hudson

0.16

æĥ³åĪ°

0.15

 encounter

0.15

ÃĮ

0.15

 encountered

0.14

 learned

0.14

 Likes

0.14

anto

0.14

 recommending

0.14

 recommend

0.14

Activations Density 0.088%

expressions of personal opinions and intentions

No Comments

No Known Activations