INDEX

Explanations

phrases that include the word "note" or "that."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

nett

-0.16

arters

-0.14

ye

-0.14

XB

-0.14

rag

-0.14

 Hacker

-0.14

nant

-0.14

ucht

-0.13

orate

-0.13

ickers

-0.13

POSITIVE LOGITS

uyen

0.15

 unlike

0.15

quam

0.14

ay

0.14

å½

0.14

æŀ

0.14

ikel

0.14

ENCHMARK

0.13

ATEGORY

0.13

ERY

0.13

Activations Density 0.038%

phrases that include the word "note" or "that."

No Comments

No Known Activations