INDEX

Explanations

names and identifiers related to people, places, and items in lists or categorization contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

aÄĩ

-0.14

ancel

-0.14

 rival

-0.14

mage

-0.14

phia

-0.13

uffman

-0.13

burgh

-0.13

 various

-0.13

esor

-0.12

ãģĿãģĨ

-0.12

POSITIVE LOGITS

 excluded

0.17

 only

0.16

only

0.16

ustum

0.15

 overlaps

0.15

 ONLY

0.15

pig

0.15

 INCLUDED

0.15

ono

0.14

Pig

0.14

Activations Density 0.158%

names and identifiers related to people, places, and items in lists or categorization contexts

No Comments

No Known Activations