INDEX

Explanations

references to contextual dependency or relational concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

uten

-0.15

ety

-0.15

 Kits

-0.15

 haze

-0.14

mods

-0.14

avl

-0.14

HS

-0.14

hoff

-0.14

readcr

-0.14

IRO

-0.14

POSITIVE LOGITS

DidLoad

0.16

ouz

0.15

antro

0.14

folio

0.14

á»ķ

0.14

lients

0.14

 NDEBUG

0.14

krom

0.14

 Beste

0.14

 Fres

0.13

Activations Density 0.001%

references to contextual dependency or relational concepts

No Comments

No Known Activations