INDEX

Explanations

actions, states, or concepts that indicate progress, change, or evaluation

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

 cref

-0.16

MeasureSpec

-0.13

rvÃ©

-0.13

 addCriterion

-0.12

 trÃº

-0.12

Ø¯ÛĮÚ¯Ø±

-0.12

Ä±rak

-0.11

[`

-0.11

mua

-0.11

 ÑĤÐ¸Ð¿Ñĥ

-0.11

POSITIVE LOGITS

regor

0.14

mdb

0.14

lij

0.13

andalone

0.13

luet

0.13

roker

0.12

aland

0.12

abr

0.12

iore

0.12

icone

0.11

Activations Density 0.032%

actions, states, or concepts that indicate progress, change, or evaluation

No Comments

No Known Activations