INDEX

Explanations

phrases indicating popularity or recognition of individuals and works

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ojÃŃ

-0.15

à¸ģà¸ķ

-0.14

 TMPro

-0.13

anya

-0.13

cff

-0.13

oki

-0.13

ÅĻi

-0.13

aram

-0.13

adena

-0.13

/cli

-0.13

POSITIVE LOGITS

 know

0.81

 knows

0.66

 known

0.63

 Know

0.62

know

0.58

-known

0.54

Know

0.53

çŁ¥

0.53

known

0.52

-know

0.51

Activations Density 0.168%

phrases indicating popularity or recognition of individuals and works

No Comments

No Known Activations

phrases indicating popularity or recognition of individuals and works

No Comments

No Known Activations