INDEX

Explanations

patterns of unusual or unexpected text formatting and symbols

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

abra

-0.16

alendar

-0.15

Runnable

-0.14

oto

-0.14

imet

-0.13

urse

-0.13

_GRAPH

-0.13

illing

-0.13

 Thing

-0.13

ÑħÐ¾Ð²

-0.13

POSITIVE LOGITS

 safety

0.17

 Safety

0.15

 safe

0.15

 yourself

0.14

 MOST

0.14

-fontawesome

0.14

teri

0.14

 most

0.14

Safety

0.14

oya

0.14

Activations Density 0.006%

patterns of unusual or unexpected text formatting and symbols

No Comments

No Known Activations