INDEX

Explanations

references to satire and satirical content.

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

quier

-0.19

_PATCH

-0.18

slt

-0.17

acl

-0.16

ktor

-0.15

erable

-0.15

ftar

-0.15

iola

-0.15

nothrow

-0.15

Å¾e

-0.14

POSITIVE LOGITS

irical

0.35

anic

0.31

ellite

0.30

suma

0.29

ellites

0.28

elite

0.28

uration

0.28

rapy

0.25

ires

0.23

iation

0.23

Activations Density 0.010%

references to satire and satirical content.

No Comments

No Known Activations