INDEX

Explanations

percentages and guarantees related to safety or quality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

pod

-0.16

ooter

-0.15

olon

-0.15

ard

-0.15

aw

-0.14

ctx

-0.14

-0.13

jen

-0.13

athan

-0.13

POSITIVE LOGITS

/full

0.18

adele

0.18

edly

0.15

portun

0.15

ometr

0.14

eru

0.14

egie

0.14

icer

0.14

 modne

0.14

\/\/

0.13

Activations Density 0.017%

percentages and guarantees related to safety or quality

No Comments

No Known Activations