INDEX

Explanations

terms associated with actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

gone

-0.15

ÏģÎµÏħ

-0.14

erot

-0.14

uve

-0.14

INTR

-0.14

alia

-0.14

alto

-0.14

gay

-0.13

ounty

-0.13

stin

-0.13

POSITIVE LOGITS

0.15

onet

0.15

tures

0.14

gnore

0.14

nings

0.14

getc

0.14

clar

0.14

oles

0.14

/out

0.14

 Ange

0.14

Activations Density 0.221%

terms associated with actions or processes

No Comments

No Known Activations