INDEX

Explanations

negations and forms of the verb "to be."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

ector

-0.14

 ÐĴÑĸÐº

-0.14

 Slate

-0.14

 Bret

-0.14

inds

-0.14

ourke

-0.14

agnost

-0.14

 pocket

-0.13

 Pocket

-0.13

bug

-0.13

POSITIVE LOGITS

justify

0.15

bane

0.14

 justify

0.14

ÙĤÙĬ

0.14

 Geile

0.14

mil

0.14

UInteger

0.14

mrb

0.14

seo

0.13

Activations Density 0.001%

negations and forms of the verb "to be."

No Comments

No Known Activations