INDEX

Explanations

references to eligibility or requirements for different situations

oai_token-act-pair · gpt-3.5-turbo

terms related to eligibility

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

ĸļ

-0.87

sie

-0.77

 constant

-0.71

 Bild

-0.67

gang

-0.66

ipes

-0.64

ience

-0.63

boys

-0.63

anoia

-0.63

velength

-0.62

POSITIVE LOGITS

 eligible

3.65

eligible

2.67

 ineligible

2.63

 elig

2.15

 eligibility

1.97

igible

1.62

 qualifying

1.55

 qualified

1.53

 enrolled

1.52

 qualify

1.48

Activations Density 0.015%

references to eligibility or requirements for different situations

terms related to eligibility

No Comments

No Known Activations