INDEX

Explanations

references to religious or historical figures and concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

eros

-0.18

ipl

-0.17

plorer

-0.15

_managed

-0.14

sell

-0.14

wnd

-0.14

ç¿¼

-0.14

 Child

-0.14

 Hatch

-0.14

Lod

-0.14

POSITIVE LOGITS

putc

0.16

æģ¨

0.15

olley

0.14

pret

0.14

RowAt

0.14

isk

0.13

Ð¾Ð·Ð°

0.13

ÏĥÎº

0.13

 Giov

0.13

é¦Ļ

0.13

Activations Density 0.092%

references to religious or historical figures and concepts

No Comments

No Known Activations