INDEX

Explanations

references to physical injuries and disabilities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

etary

-0.16

 bruises

-0.16

outers

-0.15

 Huffman

-0.14

undry

-0.14

åīĤ

-0.14

 needle

-0.13

quares

-0.13

 Shield

-0.13

needle

-0.13

POSITIVE LOGITS

 prost

0.36

 limb

0.35

 stump

0.32

 limbs

0.30

lim

0.25

Lim

0.25

am

0.23

 socket

0.21

 severed

0.21

 artificial

0.21

Activations Density 0.041%

references to physical injuries and disabilities

No Comments

No Known Activations