INDEX

Explanations

numerical representations, likely related to ratings or scores

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

DRV

-0.16

é¦

-0.15

olet

-0.15

Ø§ÙĩÛĮ

-0.15

 addCriterion

-0.15

ÑĢÑĥÐ´

-0.15

ffset

-0.14

ÃŁen

-0.14

ÏģÎ¿Ïį

-0.14

itizer

-0.14

POSITIVE LOGITS

å

0.16

Wed

0.15

AllWindows

0.15

ForObject

0.14

 replica

0.14

½

0.14

uiten

0.14

 Ø§ÙĪØª

0.14

 spread

0.14

DES

0.14

Activations Density 0.006%

numerical representations, likely related to ratings or scores

No Comments

No Known Activations