INDEX

Explanations

references to "gore," particularly in relation to its violent or horrific context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

olib

-0.16

ä¸Ńæĸĩ

-0.15

ATTERN

-0.14

coop

-0.14

ÙĥØ©

-0.14

wiki

-0.14

rint

-0.14

nP

-0.13

 lick

-0.13

\Active

-0.13

POSITIVE LOGITS

 below

0.16

 public

0.15

Tu

0.15

pray

0.14

_UNUSED

0.14

 McInt

0.14

 fract

0.14

agraph

0.14

fee

0.13

isty

0.13

Activations Density 0.003%

references to "gore," particularly in relation to its violent or horrific context

No Comments

No Known Activations