INDEX

Explanations

phrases indicating a high probability or potential outcome

oai_token-act-pair · gpt-3.5-turbo

expressions of probability or likelihood

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Columb

-0.68

-0.67

feeding

-0.64

ById

-0.63

CLASSIFIED

-0.63

raq

-0.62

iful

-0.62

Trivia

-0.61

raint

-0.61

 Palest

-0.60

POSITIVE LOGITS

be

1.16

 prove

1.02

 become

1.02

 explode

1.00

 suffice

0.98

 succeed

0.96

 revert

0.94

 qualify

0.94

 prevail

0.94

 accompany

0.94

Activations Density 0.077%

phrases indicating a high probability or potential outcome

expressions of probability or likelihood

No Comments

No Known Activations