INDEX

Explanations

phrases indicating improvement or resolution of issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.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

 OVERRIDE

-0.14

ont

-0.14

intr

-0.13

wen

-0.13

bytes

-0.13

gon

-0.13

veled

-0.13

ÙĲÙĬ

-0.13

olic

-0.13

Ì

-0.13

POSITIVE LOGITS

"url

0.16

culate

0.15

GetMethod

0.15

enaire

0.14

EMPLARY

0.14

uur

0.14

 Erotische

0.14

oÅĻ

0.14

nid

0.14

./(

0.13

Activations Density 0.882%

phrases indicating improvement or resolution of issues

No Comments

No Known Activations