INDEX

Explanations

numerical data or percentages mentioned in a document

oai_token-act-pair · gpt-3.5-turbo

sentence-ending punctuation marks and their associated numerical or percentage references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 orche

-0.87

 hostage

-0.83

 disemb

-0.77

 discreet

-0.77

 agent

-0.77

 guise

-0.76

 inspiration

-0.75

 instinct

-0.75

 defe

-0.74

 abduct

-0.74

POSITIVE LOGITS

 Similarly

1.45

 Conversely

1.45

 Overall

1.37

 Additionally

1.35

 Meanwhile

1.34

 Compared

1.34

 Among

1.30

 Likewise

1.29

 Furthermore

1.28

 Notably

1.26

Activations Density 0.334%

numerical data or percentages mentioned in a document

sentence-ending punctuation marks and their associated numerical or percentage references

No Comments

No Known Activations