INDEX

Explanations

references to specific locations, particularly ones with the word "Garden"

oai_token-act-pair · gpt-3.5-turbo

references to the term "Garden."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

senal

-0.79

 derail

-0.77

aundering

-0.76

pez

-0.72

esty

-0.72

etimes

-0.72

merce

-0.71

informed

-0.69

ilitary

-0.69

ickr

-0.69

POSITIVE LOGITS

 Garden

1.36

keeper

0.94

 Grove

0.90

 Eden

0.90

éļ

0.84

 Bowl

0.84

 Gardens

0.83

hire

0.81

 Mansion

0.80

esville

0.79

Activations Density 0.007%

references to specific locations, particularly ones with the word "Garden"

references to the term "Garden."

No Comments

No Known Activations