INDEX

Explanations

instances of the word "correct" and related expressions indicating accuracy or validity

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

udd

-0.18

/desktop

-0.17

icap

-0.15

loub

-0.15

PHY

-0.15

holding

-0.15

olt

-0.15

Ŀ

-0.15

ropping

-0.15

ary

-0.15

POSITIVE LOGITS

zza

0.21

ives

0.18

IVES

0.16

itude

0.15

mente

0.15

fully

0.15

backs

0.14

respond

0.14

_runtime

0.14

 ChÃºa

0.14

Activations Density 0.030%

instances of the word "correct" and related expressions indicating accuracy or validity

No Comments

No Known Activations