INDEX

Explanations

descriptors related to actions and interactions, specifically those that pertain to sequences and conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

ento

-0.17

ÑĥÐ¼Ñĥ

-0.16

obra

-0.16

linger

-0.16

 Mineral

-0.15

tos

-0.15

Å¡it

-0.14

xes

-0.14

ONO

-0.14

POSITIVE LOGITS

ukt

0.16

iser

0.16

legg

0.15

orial

0.14

RG

0.14

 Prev

0.14

noch

0.14

lette

0.14

ox

0.14

Activations Density 0.172%

descriptors related to actions and interactions, specifically those that pertain to sequences and conditions

No Comments

No Known Activations