INDEX

Explanations

phrases indicating protection or safety from various dangers or negative influences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 19-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L19R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.19.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

 sparing

-0.14

acific

-0.14

<<<

-0.14

èĥ¶

-0.14

ahl

-0.14

ĶĶ

-0.14

/fwlink

-0.13

luent

-0.13

 ÙħØ´Ú©

-0.13

oyer

-0.13

POSITIVE LOGITS

 scrutiny

0.25

 attack

0.24

 being

0.20

 criticism

0.19

 becoming

0.19

 harm

0.18

 harms

0.18

 attacks

0.18

æĿ¥èĩª

0.18

 further

0.17

Activations Density 0.160%

phrases indicating protection or safety from various dangers or negative influences

No Comments

No Known Activations