INDEX

Explanations

complex constructs and patterns in the text, including sentences and phrases in a specific language

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

 spoken

-0.15

 saber

-0.15

 Motion

-0.14

 Near

-0.14

CLS

-0.14

çļĦæĺ¯

-0.14

 Pier

-0.14

ed

-0.14

 bitten

-0.14

 ÑĢÐ°ÑģÐ¿ÑĢÐ¾ÑģÑĤÑĢÐ°Ð½

-0.14

POSITIVE LOGITS

 undert

0.20

got

0.19

took

0.17

ftware

0.17

 took

0.17

aday

0.16

 broke

0.15

cus

0.15

ÐµÑĢÐº

0.14

ROTO

0.14

Activations Density 0.032%

complex constructs and patterns in the text, including sentences and phrases in a specific language

No Comments

No Known Activations