INDEX

Explanations

reported speech and quotations in text

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

chl

-0.18

evin

-0.17

ics

-0.17

alte

-0.16

ongyang

-0.15

Ø±Ùĩ

-0.15

olet

-0.15

orsi

-0.15

klad

-0.15

audi

-0.15

POSITIVE LOGITS

 goodbye

0.23

/do

0.21

/write

0.18

 repeatedly

0.17

 hello

0.17

urai

0.16

ings

0.15

 despre

0.15

 nothing

0.14

-sex

0.14

Activations Density 0.197%

reported speech and quotations in text

No Comments

No Known Activations