INDEX

Explanations

terms related to excessive web content or filler text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

onaut

-0.14

wik

-0.14

idot

-0.14

ubbo

-0.14

postData

-0.14

atitude

-0.14

olib

-0.14

iesel

-0.14

 gros

-0.13

aty

-0.13

POSITIVE LOGITS

urum

0.15

 Sentry

0.14

roe

0.14

yme

0.14

Ð¾ÑģÐº

0.14

entes

0.14

ENA

0.13

é£¾

0.13

.IContainer

0.13

uctor

0.13

Activations Density 0.002%

terms related to excessive web content or filler text

No Comments

No Known Activations