INDEX

Explanations

key terms or phrases related to various subjects, especially in a contextual or directive manner

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

utto

-0.18

land

-0.16

ToDevice

-0.15

ermen

-0.15

Ø§ÙĬØ§Øª

-0.15

Ð¾ÑĢÐ´Ð¸Ð½

-0.15

atha

-0.14

è¯Ħä»·

-0.14

à¸¢à¸¡

-0.14

idor

-0.14

POSITIVE LOGITS

acon

0.19

aced

0.16

AC

0.16

dac

0.15

elic

0.15

adam

0.15

actable

0.15

 dash

0.14

umph

0.14

Bra

0.14

Activations Density 0.044%

key terms or phrases related to various subjects, especially in a contextual or directive manner

No Comments

No Known Activations