INDEX

Explanations

references to live presenting and broadcasting activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

gili

-0.20

.scalablytyped

-0.18

pane

-0.17

ylv

-0.16

AndView

-0.15

èī

-0.15

GuidId

-0.14

aru

-0.14

eres

-0.14

eh

-0.13

POSITIVE LOGITS

 prom

0.23

 talent

0.23

GFX

0.21

 Talent

0.21

tal

0.20

chy

0.20

 rund

0.19

ENG

0.19

ÂłPS

0.19

 liners

0.19

Activations Density 0.026%

references to live presenting and broadcasting activities

No Comments

No Known Activations