INDEX

Explanations

repeated use of the word "of" and related high-frequency lexical items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

INGER

-0.15

Argb

-0.14

ÙħØ±

-0.14

inger

-0.14

rams

-0.14

EXEC

-0.14

ÎºÏĮ

-0.14

laÅŁ

-0.14

hang

-0.14

iare

-0.14

POSITIVE LOGITS

neys

0.15

ãĥĸãĥ©

0.14

 solic

0.14

 Norm

0.14

isci

0.14

ammer

0.14

OA

0.14

 Bits

0.13

ella

0.13

Activations Density 0.003%

repeated use of the word "of" and related high-frequency lexical items

No Comments

No Known Activations