INDEX

Explanations

phrases related to instructions and recommendations for action

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

ittal

-0.17

ivol

-0.16

ostel

-0.16

ahoma

-0.16

zens

-0.15

.intellij

-0.15

ignon

-0.15

roperty

-0.15

oser

-0.14

reon

-0.14

POSITIVE LOGITS

 chances

0.19

inis

0.15

avage

0.15

rama

0.15

ather

0.15

 Casc

0.14

ä¾¿

0.14

IAS

0.14

é³

0.14

din

0.14

Activations Density 0.107%

phrases related to instructions and recommendations for action

No Comments

No Known Activations