INDEX

Explanations

references to specific locations, particularly names and addresses

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

Not in Any Lists

No Comments

Negative Logits

å¸¯

-0.15

zman

-0.15

("'"

-0.15

 Lehr

-0.14

ni

-0.14

avras

-0.14

glich

-0.14

 hydr

-0.14

entlich

-0.14

liv

-0.14

POSITIVE LOGITS

bach

0.29

egg

0.24

ried

0.22

acker

0.22

wang

0.22

etal

0.21

ental

0.21

br

0.20

atal

0.20

berg

0.20

Activations Density 0.046%

references to specific locations, particularly names and addresses

No Comments

No Known Activations