INDEX

Explanations

mentions of pretzel

oai_token-act-pair · gpt-3.5-turbo

references to the word "pret" and its variations

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

SHIP

-0.74

-0.73

stocks

-0.70

inventoryQuantity

-0.69

 GOODMAN

-0.67

[|

-0.66

OTOS

-0.65

 Annotations

-0.64

 Carbuncle

-0.60

hiro

-0.60

POSITIVE LOGITS

entious

1.19

zel

1.10

ensions

1.09

ension

1.04

reatment

1.03

inen

0.96

reated

0.91

ense

0.89

reating

0.89

rial

0.89

Activations Density 0.013%

mentions of pretzel

references to the word "pret" and its variations

No Comments

No Known Activations