INDEX

Explanations

certain verbs that indicate assertion or emphasis, particularly linked to states of being or conditionals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

itos

-0.15

.metro

-0.14

Pry

-0.14

ENTA

-0.14

å¤

-0.14

rompt

-0.14

illisecond

-0.13

Ù¾Ø³

-0.13

zza

-0.13

roma

-0.13

POSITIVE LOGITS

rix

0.17

ADX

0.16

Äįan

0.15

Äįin

0.15

crest

0.14

æ¡ĳ

0.14

ACHE

0.14

 easier

0.14

.DEFINE

0.13

CRE

0.13

Activations Density 0.188%

certain verbs that indicate assertion or emphasis, particularly linked to states of being or conditionals

No Comments

No Known Activations