INDEX

Explanations

phrases related to importance or impact

oai_token-act-pair · gpt-3.5-turbo

terms related to significance and influence

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

isa

-0.69

alone

-0.67

ifa

-0.67

yne

-0.64

 sails

-0.64

gets

-0.63

usters

-0.61

ãĤ´ãĥ³

-0.61

sbm

-0.60

ETA

-0.60

POSITIVE LOGITS

 dangers

0.79

 inherent

0.77

 extent

0.76

 role

0.74

 evolution

0.72

 hazards

0.71

 Jehovah

0.70

 implications

0.69

 ramifications

0.69

 usefulness

0.67

Activations Density 0.271%

phrases related to importance or impact

terms related to significance and influence

No Comments

No Known Activations