INDEX

Explanations

mentions of the location "Victoria"

oai_token-act-pair · gpt-3.5-turbo

the name "Victoria" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

earance

-0.89

mble

-0.88

backer

-0.84

bler

-0.80

kefeller

-0.79

dk

-0.77

vous

-0.77

SpaceEngineers

-0.75

respons

-0.75

linger

-0.73

POSITIVE LOGITS

 Beckham

1.01

 Falls

0.80

 Victoria

0.77

 Police

0.76

 Post

0.74

 Fleming

0.74

 Broadcasting

0.74

 Pend

0.72

 Gardens

0.72

 Park

0.71

Activations Density 0.007%

mentions of the location "Victoria"

the name "Victoria" in various contexts

No Comments

No Known Activations