INDEX

Explanations

technical code snippets indicating an increase or decrease in value

oai_token-act-pair · gpt-3.5-turbo

symbols and punctuation marks that denote emphasis or significance

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

pit

-0.80

 cabbage

-0.75

hug

-0.72

 idol

-0.72

 swim

-0.71

 mull

-0.69

 welcome

-0.68

 feared

-0.68

 paddle

-0.67

 quadru

-0.67

POSITIVE LOGITS

blogspot

0.96

olson

0.96

 Whilst

0.94

 Otherwise

0.94

 Alternatively

0.93

 Afterwards

0.91

wav

0.91

wordpress

0.90

 Presumably

0.89

 This

0.88

Activations Density 0.035%

technical code snippets indicating an increase or decrease in value

symbols and punctuation marks that denote emphasis or significance

No Comments

No Known Activations