INDEX

Explanations

terms related to tracking progress and direction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

imin

-0.16

 VerfÃ¼g

-0.16

-0.14

 paradox

-0.14

 reve

-0.14

Ã¼cken

-0.13

éľŀ

-0.13

ione

-0.13

ob

-0.13

raw

-0.13

POSITIVE LOGITS

track

0.25

 track

0.25

 direction

0.22

 Track

0.22

 course

0.22

_track

0.21

-track

0.20

course

0.20

 tracks

0.20

.track

0.19

Activations Density 0.094%

terms related to tracking progress and direction

No Comments

No Known Activations