INDEX

Explanations

negatively connotated words and phrases

oai_token-act-pair · gpt-3.5-turbo

phrases related to competitive scenarios and sports

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ĵĺ

-0.84

OST

-0.74

RAW

-0.74

APS

-0.70

olls

-0.69

 nominate

-0.67

ĳ

-0.65

actionDate

-0.64

Letter

-0.64

pora

-0.63

POSITIVE LOGITS

 hybrid

0.96

 syndrome

0.95

 hybrids

0.94

 mentality

0.87

'-

0.85

ned

0.79

 extraord

0.79

-'

0.79

 goodness

0.76

Activations Density 0.297%

negatively connotated words and phrases

phrases related to competitive scenarios and sports

No Comments

No Known Activations