INDEX

Explanations

words and phrases that indicate correctness or adherence to standards

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

-0.20

usz

-0.15

ÏģÏĮ

-0.15

-generator

-0.15

acters

-0.15

voy

-0.14

icap

-0.14

arine

-0.14

ãĥŃãĥ¼

-0.14

ìĭ

-0.14

POSITIVE LOGITS

fully

0.21

ately

0.18

 latter

0.16

getManager

0.16

FULL

0.15

afen

0.15

cast

0.15

ape

0.15

full

0.14

ourke

0.14

Activations Density 0.044%

words and phrases that indicate correctness or adherence to standards

No Comments

No Known Activations