INDEX

Explanations

architectural terms and references

oai_token-act-pair · gpt-3.5-turbo

terms related to architecture and architectural professionals

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

awar

-0.73

zees

-0.72

nir

-0.72

othal

-0.70

TPS

-0.70

eworthy

-0.70

wagon

-0.69

FFFF

-0.69

mma

-0.67

ting

-0.66

POSITIVE LOGITS

urally

1.54

itect

1.41

ural

1.21

 architect

1.05

 architecture

0.92

 architectural

0.89

 Architecture

0.88

Arch

0.87

 Architect

0.85

 architects

0.81

Activations Density 0.042%

architectural terms and references

terms related to architecture and architectural professionals

No Comments

No Known Activations