INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

terms related to hospitality and service industries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

=-=-=-=-=-=-=-=-

-0.87

atform

-0.79

rift

-0.79

ikuman

-0.78

busters

-0.77

geries

-0.74

alore

-0.72

izoph

-0.70

inates

-0.70

acters

-0.69

POSITIVE LOGITS

 hospitality

0.91

DAQ

0.81

giving

0.75

 offic

0.73

ndra

0.70

Pwr

0.68

itual

0.65

 externalToEVAOnly

0.65

osponsors

0.65

 hosp

0.64

Activations Density 0.026%

proper nouns

terms related to hospitality and service industries

No Comments

No Known Activations