INDEX

Explanations

periods and ellipses that indicate significant pauses or breaks in text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

 ÑģÐ¾ÑĢ

-0.16

ite

-0.16

alach

-0.15

aring

-0.14

 handguns

-0.13

achuset

-0.13

RITE

-0.13

 rebell

-0.13

 whipping

-0.13

POSITIVE LOGITS

ghest

0.16

eprom

0.15

BB

0.15

vas

0.15

EDITOR

0.15

aset

0.15

 Pred

0.15

 Sext

0.14

Pred

0.14

ysl

0.14

Activations Density 0.329%

periods and ellipses that indicate significant pauses or breaks in text

No Comments

No Known Activations