INDEX

Explanations

mentions of specific individuals and their affiliations or characteristics

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

sample

-0.77

 discrep

-0.76

runtime

-0.76

 situational

-0.76

norm

-0.73

reading

-0.72

response

-0.72

Sensor

-0.72

Reviewed

-0.69

 readings

-0.69

POSITIVE LOGITS

 shareholders

1.45

 shareholder

1.34

 investor

1.29

 Investments

1.27

 millionaire

1.26

 royalties

1.26

 philanthrop

1.24

 profits

1.23

 billionaire

1.23

 owns

1.23

Activations Density 0.757%

mentions of specific individuals and their affiliations or characteristics

No Comments

No Known Activations