INDEX

Explanations

second-person pronouns and phrases that suggest action or opportunities for the reader

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

inspace

-0.16

bsp

-0.16

 Mayer

-0.15

jax

-0.15

bserv

-0.15

osi

-0.14

ì§Ŀ

-0.14

istar

-0.14

Ð±ÑĥÐ´ÑĮ

-0.14

estroy

-0.14

POSITIVE LOGITS

cko

0.19

ipt

0.17

 sure

0.17

 truly

0.16

Spo

0.16

 surely

0.16

_defs

0.15

can

0.15

 certainly

0.15

 spoiled

0.15

Activations Density 0.066%

second-person pronouns and phrases that suggest action or opportunities for the reader

No Comments

No Known Activations