INDEX

Explanations

terms related to magical or extraordinary concepts

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

 Einsatz

-0.14

chen

-0.14

atisfaction

-0.14

eling

-0.13

ourcing

-0.13

Creators

-0.13

egade

-0.13

Ø§Ø°

-0.13

 Franc

-0.13

POSITIVE LOGITS

ibel

0.17

onto

0.15

ofile

0.15

á»ĳt

0.15

ibri

0.15

 unto

0.14

SZ

0.14

.fac

0.13

ãĤº

0.13

Moj

0.13

Activations Density 0.551%

terms related to magical or extraordinary concepts

No Comments

No Known Activations