INDEX

Explanations

possessive pronouns and nouns related to individuals, especially in legal and personal contexts

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

kop

-0.16

 fisse

-0.14

 æħ

-0.14

ÙĪØ³ÛĮ

-0.14

ÐµÑĢÐµÐ´

-0.14

ĮĴ

-0.14

 Candidates

-0.14

/INFO

-0.14

 æļ

-0.14

POSITIVE LOGITS

 case

0.25

 situation

0.23

 story

0.22

 condition

0.21

 predic

0.20

 whereabouts

0.20

 remains

0.19

 status

0.19

 plight

0.19

 lawyers

0.18

Activations Density 0.246%

possessive pronouns and nouns related to individuals, especially in legal and personal contexts

No Comments

No Known Activations