INDEX

Explanations

keywords related to bases or foundations

oai_token-act-pair · gpt-3.5-turbo

references to foundational elements or levels in various contexts

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

selves

-0.82

 Dragonbound

-0.76

icut

-0.73

odore

-0.73

sen

-0.72

inness

-0.72

soDeliveryDate

-0.70

odd

-0.68

alam

-0.67

byter

-0.67

POSITIVE LOGITS

 base

1.06

camp

0.87

base

0.86

 camp

0.77

 bases

0.76

plate

0.74

 Base

0.71

 garrison

0.70

 dwelling

0.70

 structure

0.68

Activations Density 0.010%

keywords related to bases or foundations

references to foundational elements or levels in various contexts

No Comments

No Known Activations