INDEX

Explanations

adjectives and descriptors that indicate specificity, accuracy, and emphasis on detail

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

arc

-0.16

WXYZ

-0.15

 enough

-0.15

ongan

-0.15

ved

-0.14

ç³»

-0.14

oup

-0.14

aku

-0.14

ane

-0.14

ooks

-0.14

POSITIVE LOGITS

-than

0.38

 than

0.36

than

0.32

Than

0.30

_than

0.29

 THAN

0.28

 Than

0.28

_THAN

0.24

 niÅ¼

0.24

 neÅ¾

0.22

Activations Density 0.115%

adjectives and descriptors that indicate specificity, accuracy, and emphasis on detail

No Comments

No Known Activations