INDEX

Explanations

words related to realms or domains, possibly related to legislation or technology

oai_token-act-pair · gpt-3.5-turbo

terms related to critical components and connections in a system

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

======

-0.75

Naz

-0.70

Bei

-0.68

========

-0.66

ļé

-0.65

============

-0.64

CE

-0.61

================

-0.61

====

-0.61

 Mats

-0.57

POSITIVE LOGITS

ether

1.13

theless

0.99

soever

0.92

landish

0.88

hemy

0.86

spr

0.86

rack

0.86

ionage

0.85

itional

0.84

icut

0.81

Activations Density 0.006%

words related to realms or domains, possibly related to legislation or technology

terms related to critical components and connections in a system

No Comments

No Known Activations