INDEX

Explanations

proper nouns, particularly names and email addresses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

loom

-0.17

urn

-0.17

elix

-0.15

æĬĺ

-0.15

ãģ³

-0.15

ava

-0.14

lish

-0.14

cher

-0.14

ello

-0.14

POSITIVE LOGITS

 Äĳ

0.15

II

0.15

oeff

0.14

ObjectType

0.14

RowIndex

0.14

 íĴĢ

0.14

 Parkway

0.13

fragistics

0.13

FixedSize

0.13

ALIGN

0.13

Activations Density 0.008%

proper nouns, particularly names and email addresses

No Comments

No Known Activations