INDEX

Explanations

references to feelings of shame and associated concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 3-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L3R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.3.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

esser

-0.16

ledo

-0.15

zos

-0.14

iets

-0.14

 Fritz

-0.14

erno

-0.14

vertime

-0.13

icut

-0.13

ockets

-0.13

POSITIVE LOGITS

lessly

0.21

fully

0.21

ishly

0.16

ously

0.16

addock

0.15

ulen

0.15

.cx

0.15

LOCKS

0.15

broken

0.15

ÑģÐ»Ð¾Ð²

0.15

Activations Density 0.012%

references to feelings of shame and associated concepts

No Comments

No Known Activations