INDEX

Explanations

conjunctions and connective phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

æĻ

-0.16

Ã©gor

-0.15

DownList

-0.15

wnd

-0.14

adium

-0.14

gue

-0.14

pii

-0.14

chter

-0.14

ilos

-0.14

rava

-0.14

POSITIVE LOGITS

ing

0.17

0.16

ingo

0.16

iesel

0.16

 others

0.15

co

0.15

ita

0.15

dek

0.15

 cent

0.15

deck

0.15

Activations Density 0.078%

conjunctions and connective phrases

No Comments

No Known Activations