INDEX

Explanations

the repeated phrase "Never" in varying contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

olik

-0.19

Ã¡ct

-0.17

ellar

-0.17

esor

-0.16

ÄħÅ¼

-0.16

ephir

-0.16

Leaks

-0.16

TRACE

-0.16

ãĤ¤ãĥī

-0.15

THREAD

-0.15

POSITIVE LOGITS

 mind

0.29

 Never

0.29

mind

0.29

 never

0.27

Never

0.26

ending

0.25

 Mind

0.25

never

0.24

 NEVER

0.23

 ever

0.21

Activations Density 0.017%

the repeated phrase "Never" in varying contexts

No Comments

No Known Activations