INDEX

Explanations

mentions of specific locations or names

oai_token-act-pair · gpt-3.5-turbo

repeated instances of specific names and terms, particularly those related to named entities or brands

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

IFF

-0.88

DPR

-0.76

esp

-0.69

ASA

-0.65

 crop

-0.64

pter

-0.64

iffs

-0.63

 jams

-0.63

iff

-0.62

 crust

-0.62

POSITIVE LOGITS

alle

2.33

Nug

1.47

Nu

1.47

ÙĦ

1.36

Doe

1.31

uana

1.22

Lu

1.09

 Louie

1.04

oleon

1.04

oxy

1.02

Activations Density 0.049%

mentions of specific locations or names

repeated instances of specific names and terms, particularly those related to named entities or brands

No Comments

No Known Activations