INDEX

Explanations

words related to negative connotations or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

eer

-0.27

ey

-0.26

es

-0.26

ek

-0.24

ep

-0.24

ene

-0.23

ef

-0.23

esian

-0.23

-0.22

ego

-0.22

POSITIVE LOGITS

iginal

0.28

ourke

0.26

hythm

0.25

ship

0.24

outines

0.23

respond

0.22

relation

0.21

hyth

0.21

rr

0.20

aptor

0.20

Activations Density 0.141%

words related to negative connotations or actions

No Comments

No Known Activations