INDEX

Explanations

references to quantification or counting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

anc

-0.17

aby

-0.16

WARE

-0.15

Anc

-0.15

Anc

-0.15

COPY

-0.14

Ware

-0.14

ucking

-0.14

PLEX

-0.14

ud

-0.14

POSITIVE LOGITS

mor

0.18

antan

0.17

ierge

0.14

mor

0.14

="{!!

0.14

MOR

0.14

iner

0.14

calar

0.14

 èµ·

0.14

pend

0.14

Activations Density 0.175%

references to quantification or counting

No Comments

No Known Activations