INDEX

Explanations

references to individuals, particularly those named William

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

_COMPAT

-0.17

entes

-0.17

ToWorld

-0.16

 Incontri

-0.16

annis

-0.15

.AWS

-0.15

zdy

-0.15

 zastup

-0.15

Ð³Ð¾ÑĤ

-0.15

nell

-0.15

POSITIVE LOGITS

 past

0.18

 Glad

0.17

 Play

0.16

Seb

0.15

 personals

0.14

war

0.14

 member

0.14

koc

0.14

ite

0.14

opia

0.14

Activations Density 0.025%

references to individuals, particularly those named William

No Comments

No Known Activations