INDEX

Explanations

mentions of the name "Ann" followed by an associated word or phrase

oai_token-act-pair · gpt-3.5-turbo

references to the name "Ann" 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 dstg

-0.78

footed

-0.78

EStreamFrame

-0.76

 dare

-0.72

ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ

-0.71

vous

-0.71

raltar

-0.70

dden

-0.70

 Mechanical

-0.69

elsius

-0.68

POSITIVE LOGITS

ihilation

1.14

unciation

1.04

ihil

1.04

ounces

1.04

abeth

1.04

abella

1.04

iston

1.02

Ann

1.02

ouncing

0.99

ounced

0.97

Activations Density 0.006%

mentions of the name "Ann" followed by an associated word or phrase

references to the name "Ann" in various contexts

No Comments

No Known Activations