INDEX

Explanations

colons and other punctuation that could signify the start of lists or important sections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

åħ

-0.15

egen

-0.14

ãĥ³ãĥĲ

-0.14

ledi

-0.14

ÏģÎ±Î½

-0.13

icious

-0.13

/cs

-0.13

 Enumeration

-0.13

ntax

-0.13

ernen

-0.13

POSITIVE LOGITS

 Observatory

0.17

croft

0.16

ÐķÐ¡

0.15

ulled

0.15

ÑĤÐ¾Ñĩ

0.15

ves

0.15

qui

0.14

carbon

0.14

Dio

0.14

ults

0.14

Activations Density 0.002%

colons and other punctuation that could signify the start of lists or important sections

No Comments

No Known Activations