INDEX

Explanations

phrases and adjectives indicating quality or type of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

uros

-0.19

hazi

-0.16

idir

-0.16

 rather

-0.14

 almost

-0.14

RATION

-0.14

almost

-0.13

éª¨

-0.13

ral

-0.13

inas

-0.13

POSITIVE LOGITS

 anymore

0.22

 coincidence

0.22

 particularly

0.21

 necessarily

0.20

 terribly

0.18

bad

0.18

 stretch

0.17

 usual

0.17

 isolated

0.17

 slam

0.16

Activations Density 0.059%

phrases and adjectives indicating quality or type of something

No Comments

No Known Activations