INDEX

Explanations

words related to benefits or advantageous outcomes

oai_token-act-pair · gpt-3.5-turbo

references to the term "benefit" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

jong

-0.72

dar

-0.71

ule

-0.69

oresc

-0.69

els

-0.64

elman

-0.64

cens

-0.64

 nicknamed

-0.62

kun

-0.62

uscript

-0.60

POSITIVE LOGITS

 benefit

3.91

benefit

2.75

 Benefit

2.62

 benefits

2.35

benef

2.01

 benefiting

1.98

 benefited

1.96

 Benefits

1.95

 advantage

1.79

Benef

1.78

Activations Density 0.017%

words related to benefits or advantageous outcomes

references to the term "benefit" in various contexts

No Comments

No Known Activations