INDEX

Explanations

words related to categories or types

oai_token-act-pair · gpt-3.5-turbo

phrases indicating categories or classifications

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

 VIDEOS

-0.74

sbm

-0.74

 Minutes

-0.70

NAS

-0.69

arks

-0.68

ults

-0.68

Phones

-0.66

 Rings

-0.65

UNCH

-0.65

CS

-0.65

POSITIVE LOGITS

 reconciliation

0.73

lier

0.70

aer

0.70

 stranger

0.69

atism

0.67

ileged

0.67

ifier

0.67

insula

0.66

bright

0.66

 whatsoever

0.66

Activations Density 0.021%

words related to categories or types

phrases indicating categories or classifications

No Comments

No Known Activations