INDEX

Explanations

instances of violence, breaches, and notable figures in reports and discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

arsing

-0.17

.labelX

-0.16

è¾¦

-0.15

IQUE

-0.14

Ð¸Ð»Ð¸

-0.14

chers

-0.14

 Birch

-0.14

keiten

-0.14

egen

-0.14

ÎµÏį

-0.13

POSITIVE LOGITS

 being

0.29

being

0.25

 Being

0.20

Being

0.19

 sendo

0.18

 essere

0.16

_basename

0.15

 therein

0.15

owell

0.14

 bÃ½t

0.14

Activations Density 0.101%

instances of violence, breaches, and notable figures in reports and discussions

No Comments

No Known Activations