INDEX

Explanations

references to a specific term or product abbreviation (PI)

oai_token-act-pair · gpt-3.5-turbo

references to the PI (presumably referring to a particular financial index or stock)

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

 emot

-0.66

 reservation

-0.65

 abst

-0.65

 coff

-0.64

 communal

-0.64

bur

-0.63

 laps

-0.62

 Yemen

-0.61

 burial

-0.60

 Entered

-0.60

POSITIVE LOGITS

PI

4.55

PI

1.77

pi

1.76

SI

1.50

Pi

1.50

PN

1.43

PLA

1.39

CI

1.38

PU

1.38

PO

1.37

Activations Density 0.009%

references to a specific term or product abbreviation (PI)

references to the PI (presumably referring to a particular financial index or stock)

No Comments

No Known Activations