INDEX

Explanations

concepts related to subtlety and hidden aspects of behavior and communication

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

eria

-0.17

rum

-0.15

pai

-0.15

ptron

-0.14

gnore

-0.14

Ð¾ÐºÐ¾Ð»

-0.14

loom

-0.14

'url

-0.14

maal

-0.14

endale

-0.14

POSITIVE LOGITS

 hidden

0.19

 behind

0.18

 subt

0.17

0.16

Pow

0.15

 Hidden

0.15

pow

0.15

-hidden

0.15

pig

0.15

 Drew

0.15

Activations Density 0.085%

concepts related to subtlety and hidden aspects of behavior and communication

No Comments

No Known Activations