INDEX

Explanations

terms related to medical recovery or convalescence

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

Negative Logits

 éĹ

-0.18

psilon

-0.17

amoto

-0.16

hit

-0.16

 odor

-0.15

ongoose

-0.14

Ø¯ÙĨ

-0.14

quine

-0.14

bage

-0.14

 Levy

-0.14

POSITIVE LOGITS

uls

0.30

ivial

0.28

ales

0.28

olutions

0.27

ulsive

0.23

 conv

0.23

icts

0.23

ective

0.23

ocation

0.23

ulsion

0.23

Activations Density 0.003%

terms related to medical recovery or convalescence

No Known Activations