INDEX

Explanations

instances of ordinal indicators or comparative phrases

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

Hop

-0.15

isis

-0.15

Hibernate

-0.15

Äįe

-0.15

faker

-0.14

 Hitch

-0.14

Ã¡c

-0.14

 Hole

-0.14

 Hover

-0.14

Helmet

-0.14

POSITIVE LOGITS

 hand

0.96

hand

0.77

-hand

0.73

 Hand

0.73

Hand

0.70

_hand

0.65

 HAND

0.65

æīĭ

0.65

.hand

0.62

HAND

0.56

Activations Density 0.129%

instances of ordinal indicators or comparative phrases

No Comments

No Known Activations