INDEX

Explanations

the word "the" in various contexts, indicating a focus on definite articles or specific references in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 15-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L15R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.15.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

-0.14

 nackte

-0.14

hair

-0.14

ermann

-0.14

ane

-0.14

oron

-0.14

EGA

-0.14

reur

-0.14

Specification

-0.14

orting

-0.13

POSITIVE LOGITS

 amount

0.21

 Abrams

0.19

 number

0.19

 extent

0.18

 manner

0.16

 degree

0.16

amount

0.16

ADVERTISEMENT

0.15

 ÙĨÙĪØ¹

0.15

Rnd

0.15

Activations Density 0.111%

the word "the" in various contexts, indicating a focus on definite articles or specific references in the text

No Comments

No Known Activations

the word "the" in various contexts, indicating a focus on definite articles or specific references in the text

No Comments

No Known Activations