INDEX

Explanations

phrases related to official classifications or designations

oai_token-act-pair · gpt-3.5-turbo

classifications related to hazardous or harmful entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 frustrations

-0.79

 tales

-0.79

 wrench

-0.71

 ratios

-0.69

 perception

-0.69

ramids

-0.68

 rhythms

-0.68

ebus

-0.68

 comparisons

-0.67

 routines

-0.66

POSITIVE LOGITS

eligible

0.98

angered

0.97

 ineligible

0.95

 exempt

0.92

 eligible

0.91

 unfit

0.88

 Category

0.85

osher

0.82

ocide

0.80

 Tradable

0.80

Activations Density 0.346%

phrases related to official classifications or designations

classifications related to hazardous or harmful entities

No Comments

No Known Activations