INDEX

Explanations

programming-related terms, particularly related to the programming language Lisp

oai_token-act-pair · gpt-3.5-turbo

references to programming languages and technology-related terms

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

-0.67

 âĢº

-0.63

++++++++++++++++

-0.63

 Everett

-0.57

beat

-0.56

 Hayden

-0.55

 Debbie

-0.54

--------------------------------

-0.54

 HuffPost

-0.54

------------------------

-0.54

POSITIVE LOGITS

 multiplication

0.62

 accessories

0.56

 lover

0.55

 initiation

0.54

 manufact

0.54

 charger

0.54

 multiplied

0.54

uliffe

0.54

 identifier

0.54

 justifies

0.53

Activations Density 0.813%

programming-related terms, particularly related to the programming language Lisp

references to programming languages and technology-related terms

No Comments

No Known Activations