INDEX

Explanations

terms related to infrastructure and copyright infringement

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

Ī

-0.16

 Mystery

-0.16

lama

-0.15

ISTER

-0.15

icz

-0.15

iesel

-0.14

ifact

-0.14

erialize

-0.14

ULO

-0.14

ICLE

-0.14

POSITIVE LOGITS

structures

0.28

structural

0.27

structure

0.26

estruct

0.25

Inf

0.25

STRUCT

0.25

Inf

0.25

 structure

0.24

Structure

0.23

structured

0.23

Activations Density 0.006%

terms related to infrastructure and copyright infringement

No Known Activations