INDEX

Explanations

phrases expressing suggestions or proposals for improvement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

indle

-0.15

 nors

-0.14

ÅĻeh

-0.14

ãĤĪãģı

-0.13

ï½¥

-0.13

ISA

-0.13

Î¼ÎŃÎ½

-0.13

CastException

-0.13

.choose

-0.13

/latest

-0.13

POSITIVE LOGITS

 maybe

0.50

 perhaps

0.46

maybe

0.44

 Maybe

0.38

perhaps

0.38

Maybe

0.36

 Perhaps

0.34

 vielleicht

0.32

Perhaps

0.31

aybe

0.30

Activations Density 0.328%

phrases expressing suggestions or proposals for improvement

No Comments

No Known Activations