INDEX

Explanations

phrases where something "applies to" someone or something

oai_token-act-pair · gpt-3.5-turbo

phrases indicating applicability or relevance

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

ovych

-0.74

©¶æ¥µ

-0.67

iHUD

-0.66

prototype

-0.66

;;;;;;;;

-0.65

ileaks

-0.63

rious

-0.62

ighters

-0.60

 Accessed

-0.60

Used

-0.59

POSITIVE LOGITS

 practise

0.85

pless

0.78

 Practice

0.77

 practice

0.71

 compensate

0.67

psy

0.67

pmwiki

0.66

othy

0.65

å§«

0.64

 srfAttach

0.64

Activations Density 0.080%

phrases where something "applies to" someone or something

phrases indicating applicability or relevance

No Comments

No Known Activations