INDEX

Explanations

instances where the word "peg" or its variations appear in the text

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "pegging" or setting a specific value or estimate

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ous

-0.89

Interstitial

-0.83

ivity

-0.83

ONSORED

-0.75

esse

-0.72

EED

-0.72

augh

-0.72

 Hague

-0.71

ESS

-0.71

ribution

-0.70

POSITIVE LOGITS

peg

1.34

pole

0.93

asus

0.88

 pegged

0.86

shit

0.80

Peg

0.79

asi

0.76

ument

0.74

atron

0.74

senal

0.70

Activations Density 0.010%

instances where the word "peg" or its variations appear in the text

references to the concept of "pegging" or setting a specific value or estimate

No Comments

No Known Activations