INDEX

Explanations

occurrences of the word "it" and variations of "is" in different contexts

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

haust

-0.15

hand

-0.14

Ùĩ

-0.14

esson

-0.14

hart

-0.13

ulture

-0.13

nbsp

-0.13

orden

-0.13

hook

-0.13

POSITIVE LOGITS

iner

0.36

unes

0.26

chy

0.24

self

0.23

inerary

0.23

zelf

0.23

/th

0.21

'll

0.19

SELF

0.19

asca

0.18

Activations Density 0.540%

occurrences of the word "it" and variations of "is" in different contexts

No Comments

No Known Activations