INDEX

Explanations

references to specific projects or franchises, particularly related to science fiction and fantasy

oai_token-act-pair · gpt-3.5-turbo

references to the "Star Trek" and "Star Wars" franchises

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

etooth

-0.84

 tremend

-0.82

 trouble

-0.78

undai

-0.73

inarily

-0.71

 defund

-0.71

 counselor

-0.70

unciation

-0.69

 reconc

-0.67

ierrez

-0.67

POSITIVE LOGITS

ãĥīãĥ©ãĤ´ãĥ³

0.83

0.81

CV

0.79

0.77

0.76

0.75

 Origins

0.74

GO

0.74

Activations Density 0.145%

references to specific projects or franchises, particularly related to science fiction and fantasy

references to the "Star Trek" and "Star Wars" franchises

No Comments

No Known Activations