INDEX

Explanations

classifications and attributes of different groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

bow

-0.15

 Pink

-0.15

upert

-0.15

ÏģÎ±ÏĤ

-0.15

Col

-0.15

 Thin

-0.14

 material

-0.14

reeze

-0.14

 handy

-0.14

reed

-0.14

POSITIVE LOGITS

íĻ©

0.19

 casts

0.17

 Ð±Ð»

0.17

 intens

0.16

 intense

0.16

 Amar

0.16

æ¿ĥ

0.16

-viol

0.15

åº§

0.15

ELLOW

0.15

Activations Density 0.022%

classifications and attributes of different groups

No Comments

No Known Activations