INDEX

Explanations

words related to rebounding or recovering from a setback

oai_token-act-pair · gpt-3.5-turbo

terms related to recovery or improvement from a negative situation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

cas

-0.83

cho

-0.73

urat

-0.72

vironments

-0.70

iop

-0.68

gemony

-0.68

66666666

-0.67

ffe

-0.66

cale

-0.65

ATA

-0.64

POSITIVE LOGITS

 rebound

1.27

 rebounds

0.86

lund

0.79

wcs

0.79

 damp

0.79

interstitial

0.78

 eject

0.77

catentry

0.74

erred

0.72

 defic

0.72

Activations Density 0.006%

words related to rebounding or recovering from a setback

terms related to recovery or improvement from a negative situation

No Comments

No Known Activations