INDEX

Explanations

intensifiers indicating a high level of quality or competency

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 15-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L15R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.15.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

pras

-0.16

KK

-0.15

ÄĻki

-0.14

ÃŃrk

-0.14

/by

-0.13

ESA

-0.13

imb

-0.13

_rq

-0.13

/go

-0.13

 lower

-0.13

POSITIVE LOGITS

acket

0.16

ana

0.16

hled

0.15

IDA

0.14

eur

0.14

bole

0.14

eness

0.14

tons

0.13

 MessageType

0.13

Ø±ÙĬÙĦ

0.13

Activations Density 0.019%

intensifiers indicating a high level of quality or competency

No Comments

No Known Activations