INDEX

Explanations

mentions of the word "the" in relation to content about laws, governance, and social issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

enden

-0.15

 offending

-0.15

rado

-0.14

Ã¦re

-0.14

arel

-0.14

ternet

-0.14

rys

-0.14

_DISPATCH

-0.14

ENTITY

-0.13

edback

-0.13

POSITIVE LOGITS

 result

0.39

 product

0.36

product

0.30

result

0.28

 Result

0.25

 subject

0.24

 stuff

0.24

 resultado

0.24

 consequence

0.24

Product

0.24

Activations Density 0.142%

mentions of the word "the" in relation to content about laws, governance, and social issues

No Comments

No Known Activations