INDEX

Explanations

references to various types of models

oai_token-act-pair · gpt-3.5-turbo

references to various types of models

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

olulu

-0.86

vernment

-0.83

ulhu

-0.78

azard

-0.71

velength

-0.69

Downloadha

-0.69

uden

-0.67

avorite

-0.66

cling

-0.65

cffff

-0.64

POSITIVE LOGITS

model

1.25

 model

1.02

 Models

1.01

 models

0.96

models

0.94

 Model

0.91

Model

0.80

urer

0.74

 Operator

0.72

ãĤ´

0.72

Activations Density 0.016%

references to various types of models

references to various types of models

No Comments

No Known Activations